Import required libraries¶

import plotly
print(plotly.__version__)
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
%matplotlib inline
import seaborn as sns
import plotly.express as px
import plotly.graph_objects as go
from sklearn.impute import SimpleImputer
# Basic packages
import numpy as np, gc
from scipy import stats; 
from scipy.stats import zscore, norm
from scipy.stats import randint as sp_randint
import matplotlib.style as style; 
style.use('fivethirtyeight')

# Models
from sklearn.model_selection import train_test_split, GridSearchCV, StratifiedKFold, cross_val_score, learning_curve
from wordcloud import WordCloud, STOPWORDS
 
# Display settings
pd.options.display.max_rows = 500
pd.options.display.max_columns = 100
pd.options.display.float_format = "{:.2f}".format
 
random_state = 42
np.random.seed(random_state)
import copy
import holidays
# Suppress warnings
import warnings; warnings.filterwarnings('ignore')

import random, re
import time

# used to supress display of warnings
import warnings

import missingno as mno

# nlp libraries
import nltk
nltk.download('punkt')
nltk.download('stopwords')
nltk.download('wordnet');

from nltk.tokenize import word_tokenize
from tqdm import tqdm
from nltk.corpus import stopwords
stop_words = stopwords.words('english')
from sklearn.feature_extraction.text import TfidfVectorizer

from nltk.stem import WordNetLemmatizer
from nltk.corpus import wordnet as wn
from wordcloud import WordCloud
import string

from collections import defaultdict
from collections import  Counter

# import holoviews as hv
# from holoviews import opts

import os;
from os import makedirs

from gensim.models import Word2Vec

# sampling methods
from sklearn.utils import resample
from imblearn.over_sampling import SMOTE

# save models
import pickle

# pre-processing methods
from sklearn.model_selection import train_test_split

# the classification models 
from sklearn.dummy import DummyClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.linear_model import RidgeClassifier
from sklearn.linear_model import Lasso
from sklearn.neighbors import KNeighborsClassifier
from sklearn.naive_bayes import GaussianNB
from sklearn.svm import SVC
from sklearn.tree import DecisionTreeClassifier

# ensemble models
from sklearn.ensemble import RandomForestClassifier
from sklearn.ensemble import StackingClassifier
from sklearn.ensemble import BaggingClassifier
from sklearn.ensemble import ExtraTreesClassifier
from sklearn.ensemble import AdaBoostClassifier
from sklearn.ensemble import GradientBoostingClassifier
import lightgbm as lgb
from lightgbm import LGBMClassifier
from xgboost import XGBClassifier

# methods and classes for evaluation
from sklearn.model_selection import cross_val_score
from sklearn.model_selection import StratifiedKFold

from sklearn import metrics
from sklearn.metrics import classification_report
from sklearn.metrics import accuracy_score, f1_score, confusion_matrix, recall_score, precision_score, roc_auc_score

# cross-validation methods
from sklearn.model_selection import KFold
from sklearn.model_selection import RepeatedStratifiedKFold
from sklearn.model_selection import GridSearchCV, RandomizedSearchCV

# feature selection methods
from sklearn.pipeline import Pipeline
from sklearn.decomposition import PCA
from sklearn.decomposition import TruncatedSVD
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
from sklearn.feature_selection import SelectKBest
from sklearn.feature_selection import f_classif
from sklearn.feature_selection import mutual_info_classif
from sklearn.feature_selection import RFE
from sklearn.feature_selection import RFECV
from sklearn.feature_extraction.text import CountVectorizer

# pre-processing methods
from sklearn.preprocessing import StandardScaler
from sklearn.preprocessing import PowerTransformer
from sklearn.preprocessing import LabelEncoder

# Deep learning libraries
from keras.utils import np_utils
from keras.utils.vis_utils import plot_model
from keras.layers import Input
from keras.layers.merge import Concatenate
from tensorflow.keras.optimizers import SGD
from tensorflow.keras.models import Sequential
from tensorflow.keras import optimizers
from keras.models import Model
from tensorflow.keras.layers import Flatten, Activation, Dense, LSTM, BatchNormalization, Embedding, Dropout, Flatten, Bidirectional, GlobalMaxPool1D
from keras.models import model_from_json
from keras.regularizers import l1, l2, l1_l2
from keras.constraints import maxnorm, min_max_norm
from keras.constraints import unit_norm
from keras.callbacks import ReduceLROnPlateau
from keras.callbacks import EarlyStopping
from keras.callbacks import ModelCheckpoint
from keras.models import model_from_json

from keras.models import load_model
from keras.wrappers.scikit_learn import KerasClassifier

# Keras pre-processing
from tensorflow.keras.preprocessing.text import Tokenizer
from tensorflow.keras.preprocessing.sequence import pad_sequences

import tensorflow as tf

5.2.2
[nltk_data] Downloading package punkt to
[nltk_data]     C:\Users\G621315\AppData\Roaming\nltk_data...
[nltk_data]   Package punkt is already up-to-date!
[nltk_data] Downloading package stopwords to
[nltk_data]     C:\Users\G621315\AppData\Roaming\nltk_data...
[nltk_data]   Package stopwords is already up-to-date!
[nltk_data] Downloading package wordnet to
[nltk_data]     C:\Users\G621315\AppData\Roaming\nltk_data...
[nltk_data]   Package wordnet is already up-to-date!

Define common fuctions¶

# Color Class to Bold and Color words while printing
class color:
    PURPLE = '\033[95m'
    CYAN = '\033[96m'
    DARKCYAN = '\033[36m'
    BLUE = '\033[94m'
    GREEN = '\033[92m'
    YELLOW = '\033[93m'
    RED = '\033[91m'
    BOLD = '\033[1m'
    UNDERLINE = '\033[4m'
    END = '\033[0m'

# function to create month variable into seasons
def month2seasons(x):
    if x in [9, 10, 11]:
        season = 'Spring'
    elif x in [12, 1, 2]:
        season = 'Summer'
    elif x in [3, 4, 5]:
        season = 'Autumn'
    elif x in [6, 7, 8]:
        season = 'Winter'
    return season

# function to plot top Stopwords
def plot_top_stopwords_barchart(text):
    stop=set(stopwords.words('english'))
    
    new= text.str.split()
    new=new.values.tolist()
    corpus=[word for i in new for word in i]
    dic=defaultdict(int)
    for word in corpus:
        if word in stop:
            dic[word]+=1
            
    top=sorted(dic.items(), key=lambda x:x[1],reverse=True)[:10] 
    x,y=zip(*top)
    plt.bar(x,y)

#function to plot the top n-grams for a text
def plot_top_ngrams_barchart(text, n=2):
    stop=set(stopwords.words('english'))

    new= text.str.split()
    new=new.values.tolist()
    corpus=[word for i in new for word in i]

    def _get_top_ngram(corpus, n=None):
        vec = CountVectorizer(ngram_range=(n, n)).fit(corpus)
        bag_of_words = vec.transform(corpus)
        sum_words = bag_of_words.sum(axis=0) 
        words_freq = [(word, sum_words[0, idx]) 
                      for word, idx in vec.vocabulary_.items()]
        words_freq =sorted(words_freq, key = lambda x: x[1], reverse=True)
        return words_freq[:10]

    top_n_bigrams=_get_top_ngram(text,n)[:10]
    x,y=map(list,zip(*top_n_bigrams))
    sns.barplot(x=y,y=x)

#Helper functions to get cleaned text column
nltk.download('averaged_perceptron_tagger')

appos = {"ain't": "am not", "aren't": "are not", "can't": "cannot", 
         "can't've": "cannot have", "'cause": "because", 
         "could've": "could have", "couldn't": "could not", 
         "couldn't've": "could not have", "didn't": "did not", 
         "doesn't": "does not", "don't": "do not", "hadn't": "had not", 
         "hadn't've": "had not have", "hasn't": "has not", 
         "haven't": "have not", "he'd": "he would", "he'd've": "he would have", 
         "he'll": "he will", "he'll've": "he will have", 
         "he's": "he is", "how'd": "how did", 
         "how'd'y": "how do you", "how'll": "how will", "how's": "how is",
         "I'd": "I would", "I'd've": "I would have", "I'll": "I will", 
         "I'll've": "I will have", "I'm": "I am", "I've": "I have", 
         "isn't": "is not", "it'd": "it would", "it'd've": "it would have", 
         "it'll": "it will", "it'll've": "it will have", "it's": "it is", 
         "let's": "let us", "ma'am": "madam", "mayn't": "may not", 
         "might've": "might have", "mightn't": "might not", 
         "mightn't've": "might not have", "must've": "must have", 
         "mustn't": "must not", "mustn't've": "must not have", 
         "needn't": "need not", "needn't've": "need not have",
         "o'clock": "of the clock", "oughtn't": "ought not", 
         "oughtn't've": "ought not have", "shan't": "shall not", 
         "sha'n't": "shall not", "shan't've": "shall not have", 
         "she'd": "she would", "she'd've": "she would have", 
         "she'll": "she will", "she'll've": "she will have",
         "she's": "she is", "should've": "should have", 
         "shouldn't": "should not", "shouldn't've": "should not have", 
         "so've": "so have", "so's": "so is", 
         "that'd": "that had", "that'd've": "that would have", 
         "that's": "that that is", "there'd": "there would", 
         "there'd've": "there would have", "there's": "there is", 
         "they'd": "they would", "they'd've": "they would have", 
         "they'll": "they will", "they'll've": "they will have", 
         "they're": "they are", "they've": "they have", 
         "to've": "to have", "wasn't": "was not", "we'd": "we would", 
         "we'd've": "we would have", "we'll": "we will", 
         "we'll've": "we will have", "we're": "we are", 
         "we've": "we have", "weren't": "were not", 
         "what'll": "what will", "what'll've": "what will have", 
         "what're": "what are", "what's": "what is", 
         "what've": "what have", "when's": "when is", 
         "when've": "when have", "where'd": "where did", 
         "where's": "where is", "where've": "where have", 
         "who'll": "who will", "who'll've": "who will have", 
         "who's": "who is", "who've": "who have", 
         "why's": "why is", "why've": "why have", "will've": "will have", 
         "won't": "will not", "won't've": "will not have",
         "would've": "would have", "wouldn't": "would not", 
         "wouldn't've": "would not have", "y'all": "you all", 
         "y'all'd": "you all would", "y'all'd've": "you all would have", 
         "y'all're": "you all are", "y'all've": "you all have", 
         "you'd": "you would", "you'd've": "you would have",
         "you'll": "you will", "you'll've": "you will have", 
         "you're": "you are", "you've": "you have"}

# Helper function to replace appos
def replace_words(description):
    cleaned_description = []
    for word in str(description).split():
        if word.lower() in appos.keys():
            cleaned_description.append(appos[word.lower()])
        else:
            cleaned_description.append(word)
    return ' '.join(cleaned_description)

# Helper function to remove punctuations
# Reference: https://www.programiz.com/python-programming/methods/string/translate
PUNCT_TO_REMOVE = '!"#$%&\'()*+,-./:;<=>?@[\\]^_`{|}~' #string.punctuation
def remove_punctuation(text):
    """function to remove the punctuation"""
    return text.translate(str.maketrans('', '', PUNCT_TO_REMOVE))

# Helper function to lemmatize
lemmatizer = WordNetLemmatizer()
def lemmatize(text):
  return ''.join([lemmatizer.lemmatize(word) for word in text])

# Helper function to remove stopwords
stoplist = set(stopwords.words('english'))
stoplist.update(('cm', 'kg', 'mr', 'wa' ,'nv', 'ore', 'da', 'pm', 'am', 'cx'))
stoplist.remove('not')
def remove_stopwords(text):
    """function to remove the stopwords"""
    return " ".join([word for word in str(text).split() if word not in stoplist])

# Helper function for wordcloud
# Reference: https://www.kaggle.com/aashita/word-clouds-of-various-shapes
def plot_wordcloud(text, mask = None, max_words = 500, max_font_size = 40, 
                   figure_size = (12, 6), title = None, title_size = 15):
    wordcloud = WordCloud(background_color = 'white', max_words = max_words,
                          random_state = 42, width = 350, height = 150, 
                          mask = mask, stopwords = stoplist, collocations = False)
    wordcloud.generate(str(text))
    
    plt.figure(figsize = figure_size)
    plt.imshow(wordcloud, interpolation = 'bilinear');
    plt.title(title, fontdict = {'size': title_size, 'color': 'black', 
                               'verticalalignment': 'bottom'})
    plt.axis('off');
    plt.tight_layout()

# Second helper function for lemmatizing
# lemmatizer = WordNetLemmatizer()
# def lem(text):
#     pos_dict = {'N': wn.NOUN, 'V': wn.VERB, 'J': wn.ADJ, 'R': wn.ADV}
#     return(' '.join([lemmatizer.lemmatize(w,pos_dict.get(t, wn.NOUN)) for w,t in nltk.pos_tag(text.split())]))

#Functions to get cleaned text column
def get_cleaned_desc(df, col_to_clean, cleaned_col_name):
      print(color.GREEN + 'Converting description to lower case' + color.END)
      df[cleaned_col_name] = df[col_to_clean].apply(lambda x : x.lower())
      
      print(color.GREEN + 'Replacing apostrophes to the standard lexicons' + color.END)
      df[cleaned_col_name] = df[cleaned_col_name].apply(lambda x : replace_words(x))
      
      print(color.GREEN + 'Removing punctuations' + color.END)
      df[cleaned_col_name] = df[cleaned_col_name].apply(lambda x: remove_punctuation(x))
      
      print(color.GREEN + 'Applying Lemmatizer' + color.END)
      df[cleaned_col_name] = df[cleaned_col_name].apply(lambda x: lemmatize(x))
      
      print(color.GREEN + 'Removing multiple spaces between words' + color.END)
      df[cleaned_col_name] = df[cleaned_col_name].apply(lambda x: re.sub(' +', ' ', x))
      
      print(color.GREEN + 'Removing stop words' + color.END)
      df[cleaned_col_name] = df[cleaned_col_name].apply(lambda x: remove_stopwords(x))

      return df

def get_cleaned_desc_wo_print(df, col_to_clean, cleaned_col_name):
    df[cleaned_col_name] = df[col_to_clean].apply(lambda x : x.lower())
    df[cleaned_col_name] = df[cleaned_col_name].apply(lambda x : replace_words(x))
    df[cleaned_col_name] = df[cleaned_col_name].apply(lambda x: remove_punctuation(x))
    df[cleaned_col_name] = df[cleaned_col_name].apply(lambda x: lemmatize(x))
    df[cleaned_col_name] = df[cleaned_col_name].apply(lambda x: re.sub(' +', ' ', x))
    df[cleaned_col_name] = df[cleaned_col_name].apply(lambda x: remove_stopwords(x))
    return df

[nltk_data] Downloading package averaged_perceptron_tagger to
[nltk_data]     C:\Users\G621315\AppData\Roaming\nltk_data...
[nltk_data]   Package averaged_perceptron_tagger is already up-to-
[nltk_data]       date!

# this function creates a normalized vector for the whole sentence
def sent2vec(s):
    words = str(s).lower()
    words = word_tokenize(words)
    words = [w for w in words if not w in stop_words]
    words = [w for w in words if w.isalpha()]
    M = []
    for w in words:
        try:
            M.append(embeddings_index[w])
        except:
            continue
    M = np.array(M)
    v = M.sum(axis=0)
    if type(v) != np.ndarray:
        return np.zeros(300)
    return v / np.sqrt((v ** 2).sum())

#Helper classes for Machine Learning, Neural Network and NLP models

def multiclass_logloss(actual, predicted, eps=1e-15):
    """Multi class version of Logarithmic Loss metric.
    :param actual: Array containing the actual target classes
    :param predicted: Matrix with class predictions, one probability per class
    """
    # Convert 'actual' to a binary array if it's not already:
    if len(actual.shape) == 1:
        actual2 = np.zeros((actual.shape[0], predicted.shape[1]))
        for i, val in enumerate(actual):
            actual2[i, val] = 1
        actual = actual2

    clip = np.clip(predicted, eps, 1 - eps)
    rows = actual.shape[0]
    vsota = np.sum(actual * np.log(clip))
    return -1.0 / rows * vsota

def train_test_model(model, method, X_train, X_test, y_train, y_test, of_type, index, scale, report, save_model):
    
    if report == "yes":
        print (model)
        print ("***************************************************************************")

    if method == 'CatBoostClassifier' or method == 'LGBMClassifier':

      model.fit(X_train, y_train) # Fit the model on Training set
    else:
      model.fit(X_train, y_train) # Fit the model on Training set

    from sklearn.metrics import accuracy_score, f1_score, confusion_matrix, recall_score, precision_score
    
    if of_type == "coef":
        # Intercept and Coefficients
        print("The intercept for our model is {}".format(model.intercept_), "\n")
        
        for idx, col_name in enumerate(X_train.columns):
            print("The coefficient for {} is {}".format(col_name, model.coef_.ravel()[idx]))

    y_pred = model.predict(X_test) # Predict on Test set

    # Initialise mc_logloss
    mc_logloss = 1.00
    if method != 'RidgeClassifier':
      y_predictions = model.predict_proba(X_test)

    train_accuracy_score = model.score(X_train, y_train)
    test_accuracy_score = model.score(X_test, y_test)

    precision_score = precision_score(y_test, y_pred, average='weighted')
    recall_score = recall_score(y_test, y_pred, average='weighted')
    f1_score = f1_score(y_test, y_pred, average='weighted')

    if method != 'RidgeClassifier':
      mc_logloss = multiclass_logloss(y_test, y_predictions, eps=1e-15)

    if report == "yes":
      # Model - Confusion matrix
      model_cm = confusion_matrix(y_test, y_pred)

      sns.heatmap(model_cm, annot=True,  fmt='.2f', xticklabels = ["I", "II", "III", "IV", "V"] , yticklabels = ["I", "II", "III", "IV", "V"] )
      plt.ylabel('Actual')
      plt.xlabel('Predicted')
      plt.show()

      # Model - Classification report
      model_cr = classification_report(y_test, y_pred)
      print(model_cr)

    # Store the accuracy results for each model in a dataframe for final comparison
    resultsDf = pd.DataFrame({'Method': method, 'Train Accuracy': train_accuracy_score, 'Test Accuracy': test_accuracy_score, 
                              'Precision': precision_score, 'Recall': recall_score, 'F1-Score': f1_score, 
                              'Multi-Class Logloss': mc_logloss}, index=[index])
    
    # Save the model
    if save_model == "yes":
      filename = 'finalised_model.sav'
      pickle.dump(model, open(filename, 'wb'))
      
    return resultsDf  # return all the metrics along with predictions
    
    
def train_test_allmodels(X_train_common, X_test_common, y_train, y_test, scale):

    # define classification models
    models=[['LogisticRegression',LogisticRegression(solver='lbfgs', multi_class='multinomial', random_state = 1)],
        ['RidgeClassifier',RidgeClassifier(random_state = 1)],
        ['KNeighborsClassifier',KNeighborsClassifier(n_neighbors = 3)],
        ['SVC',SVC(kernel = 'rbf', probability=True)],
        ['DecisionTreeClassifier',DecisionTreeClassifier(criterion = 'gini', random_state=1)],
        ['RandomForestClassifier',RandomForestClassifier(n_estimators=10, random_state=1)],
        ['BaggingClassifier',BaggingClassifier(n_estimators=30, max_samples=0.75, random_state=1, oob_score=True)],
        ['ExtraTreesClassifier',ExtraTreesClassifier(n_estimators = 50, criterion='entropy', max_features='auto', min_samples_split=2, 
                                 bootstrap=True, oob_score=True)],
        ['AdaBoostClassifier',AdaBoostClassifier(n_estimators=100, learning_rate=0.25, random_state=1)],
        ['GradientBoostingClassifier',GradientBoostingClassifier(loss='deviance', n_estimators=50, learning_rate=0.1, validation_fraction=0.2, 
                                       random_state=1)],
        ['LGBMClassifier',LGBMClassifier(random_state=1, metric = "multi_logloss", objective="multiclass")],
        ['XGBClassifier',XGBClassifier(min_child_weight = 7, max_depth = 6, objective="multi:softmax", learning_rate = 0.1, gamma = 0.4, 
                                       colsample_bytree = 0.5)]
    ]

    resultsDf_common = pd.DataFrame()
    i = 1
    for name, classifier in models:
        # Train and Test the model
        reg_resultsDf = train_test_model(classifier, name, X_train_common, X_test_common, y_train, y_test, 'none', i, scale, 'no', 'no')

        # Store the accuracy results for each model in a dataframe for final comparison
        resultsDf_common = pd.concat([resultsDf_common, reg_resultsDf])
        i = i+1

    return resultsDf_common


def hyperparameterstune_model(name, model, X_train, y_train, param_grid):
    
    start = time.time()  # note the start time 
    
    # Before starting with grid search we need to create a scoring function. This is accomplished using the make_scorer function of scikit-learn.
    mll_scorer = metrics.make_scorer(multiclass_logloss, greater_is_better=False, needs_proba=True)

    # define grid search
    cv = RepeatedStratifiedKFold(n_splits=10, n_repeats=3, random_state=1)
    if name == 'LGBMClassifier':
      grid_search = RandomizedSearchCV(estimator=model, param_distributions=param_grid, n_iter=100, n_jobs=-1, cv=cv, 
                                       scoring = mll_scorer, error_score=0)
    else:
      grid_search = GridSearchCV(estimator=model, param_grid=param_grid, n_jobs=-1, cv=cv, 
                                 scoring = mll_scorer, error_score=0)
      
    model_grid_result = grid_search.fit(X_train, y_train)

    # summarize results
    print("Best F1_Score: %f using %s" % (model_grid_result.best_score_, model_grid_result.best_params_))
    means = model_grid_result.cv_results_['mean_test_score']
    stds = model_grid_result.cv_results_['std_test_score']
    params = model_grid_result.cv_results_['params']
    for mean, stdev, param in zip(means, stds, params):
      if param == model_grid_result.best_params_:
        print("%f (%f) with: %r" % (mean, stdev, param))
        print("95% Confidence interval range: ({0:.4f} %, {1:.4f} %)".format(mean-(2*stdev), mean+(2*stdev)))

    end = time.time()  # note the end time
    duration = end - start  # calculate the total duration
    print("Total duration" , duration, "\n")
    
    return model_grid_result.best_estimator_

    
# get the accuracy, precision, recall, f1 score from model
def get_classification_metrics(model, X_test, y_test, target_type):
  
  # predict probabilities for test set
  yhat_probs = model.predict(X_test, verbose=0) # Multiclass

  # predict crisp classes for test set
  if target_type == 'multi_class':
    yhat_classes = model.predict_classes(X_test, verbose=0) # Multiclass
  else:
    yhat_classes = (np.asarray(model.predict(X_test))).round() # Multilabel

  # reduce to 1d array
  yhat_probs = yhat_probs[:, 0]

  # accuracy: (tp + tn) / (p + n)
  accuracy = accuracy_score(y_test, yhat_classes)

  # precision tp / (tp + fp)
  precision = precision_score(y_test, yhat_classes, average='micro')

  # recall: tp / (tp + fn)
  recall = recall_score(y_test, yhat_classes, average='micro')

  # f1: 2 tp / (2 tp + fp + fn)
  f1 = f1_score(y_test, yhat_classes, average='micro')

  return accuracy, precision, recall, f1


class Metrics(tf.keras.callbacks.Callback):

    def __init__(self, validation_data=()):
        super().__init__()
        self.validation_data = validation_data

    def on_train_begin(self, logs={}):
        self.val_f1s = []
        self.val_recalls = []
        self.val_precisions = []

    def on_epoch_end(self, epoch, logs={}):
        xVal, yVal, target_type = self.validation_data
        if target_type == 'multi_class':
          val_predict_classes = model.predict_classes(xVal, verbose=0) # Multiclass
        else:
          val_predict_classes = (np.asarray(self.model.predict(xVal))).round() # Multilabel
        
        
        val_targ = yVal

        _val_f1 = f1_score(val_targ, val_predict_classes, average='micro')
        _val_recall = recall_score(val_targ, val_predict_classes, average='micro')
        _val_precision = precision_score(val_targ, val_predict_classes, average='micro')
        self.val_f1s.append(_val_f1)
        self.val_recalls.append(_val_recall)
        self.val_precisions.append(_val_precision)
        #print("— train_f1: %f — train_precision: %f — train_recall %f" % (_val_f1, _val_precision, _val_recall))
        return

Import data from Google Drive¶

# from google.colab import drive

# drive.mount('/content/gdrive')

#data = pd.read_csv('/content/gdrive/MyDrive/Capstone/Data Set - industrial_safety_and_health_database_with_accidents_description.csv', index_col=None, header=0)

data = pd.read_csv('industrial_safety_and_health_database_with_accidents_description.csv', index_col=None, header=0)

print("Number of rows = " + color.GREEN + str(data.shape[0]) + color.END + " and Number of Columns = " + color.GREEN + str(data.shape[1]) + color.END + " in the Data frame")

Number of rows = 425 and Number of Columns = 11 in the Data frame

EDA and Preprocessing¶

data.dtypes

Unnamed: 0                   int64
Data                        object
Countries                   object
Local                       object
Industry Sector             object
Accident Level              object
Potential Accident Level    object
Genre                       object
Employee or Third Party     object
Critical Risk               object
Description                 object
dtype: object

data.head()

Get distinct values in each column¶

for col in data:
    if str(col)!='Description':
        print("\n" + color.BLUE + color.UNDERLINE + color.BOLD + str(col).upper() + color.END + "\n")
        print(data[col].unique())

UNNAMED: 0

[  0   1   2   3   4   5   6   7   8   9  10  11  12  13  14  15  16  17
  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35
  36  37  38  39  40  41  54  55  56  57  58  59  60  61  62  63  64  65
  66  67  68  69  70  71  72  73  74  75  76  77  78  79  80  81  82  83
  84  85  86  87  88  89  90  91  92  93  94  95  96  97  98  99 100 101
 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119
 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137
 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155
 156 157 158 159 160 161 162 163 164 165 166 169 170 171 172 173 174 175
 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193
 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211
 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229
 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247
 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283
 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301
 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319
 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337
 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355
 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373
 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391
 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409
 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427
 428 429 430 431 432 433 434 435 436 437 438]

DATA

['2016-01-01 00:00:00' '2016-01-02 00:00:00' '2016-01-06 00:00:00'
 '2016-01-08 00:00:00' '2016-01-10 00:00:00' '2016-01-12 00:00:00'
 '2016-01-16 00:00:00' '2016-01-17 00:00:00' '2016-01-19 00:00:00'
 '2016-01-26 00:00:00' '2016-01-28 00:00:00' '2016-01-30 00:00:00'
 '2016-02-01 00:00:00' '2016-02-02 00:00:00' '2016-02-04 00:00:00'
 '2016-02-06 00:00:00' '2016-02-07 00:00:00' '2016-02-08 00:00:00'
 '2016-02-21 00:00:00' '2016-02-25 00:00:00' '2016-02-09 00:00:00'
 '2016-02-10 00:00:00' '2016-02-15 00:00:00' '2016-02-14 00:00:00'
 '2016-02-13 00:00:00' '2016-02-16 00:00:00' '2016-02-17 00:00:00'
 '2016-02-19 00:00:00' '2016-02-20 00:00:00' '2016-02-18 00:00:00'
 '2016-02-22 00:00:00' '2016-02-24 00:00:00' '2016-02-29 00:00:00'
 '2016-02-26 00:00:00' '2016-02-27 00:00:00' '2016-03-02 00:00:00'
 '2016-03-03 00:00:00' '2016-03-04 00:00:00' '2016-03-05 00:00:00'
 '2016-03-06 00:00:00' '2016-03-09 00:00:00' '2016-03-11 00:00:00'
 '2016-03-13 00:00:00' '2016-03-12 00:00:00' '2016-03-14 00:00:00'
 '2016-03-16 00:00:00' '2016-03-10 00:00:00' '2016-03-17 00:00:00'
 '2016-03-18 00:00:00' '2016-03-19 00:00:00' '2016-03-22 00:00:00'
 '2016-03-25 00:00:00' '2016-03-30 00:00:00' '2016-03-31 00:00:00'
 '2016-04-01 00:00:00' '2016-04-03 00:00:00' '2016-04-02 00:00:00'
 '2016-03-24 00:00:00' '2016-04-04 00:00:00' '2016-04-05 00:00:00'
 '2016-04-07 00:00:00' '2016-04-08 00:00:00' '2016-04-11 00:00:00'
 '2016-04-14 00:00:00' '2016-04-16 00:00:00' '2016-04-15 00:00:00'
 '2016-04-17 00:00:00' '2016-04-18 00:00:00' '2016-04-21 00:00:00'
 '2016-04-22 00:00:00' '2016-04-23 00:00:00' '2016-04-26 00:00:00'
 '2016-04-28 00:00:00' '2016-04-29 00:00:00' '2016-04-30 00:00:00'
 '2016-05-01 00:00:00' '2016-05-02 00:00:00' '2016-05-04 00:00:00'
 '2016-05-03 00:00:00' '2016-05-05 00:00:00' '2016-05-11 00:00:00'
 '2016-05-12 00:00:00' '2016-05-14 00:00:00' '2016-05-17 00:00:00'
 '2016-05-19 00:00:00' '2016-05-18 00:00:00' '2016-05-22 00:00:00'
 '2016-05-20 00:00:00' '2016-05-24 00:00:00' '2016-05-25 00:00:00'
 '2016-05-27 00:00:00' '2016-05-26 00:00:00' '2016-06-01 00:00:00'
 '2016-06-02 00:00:00' '2016-06-03 00:00:00' '2016-06-04 00:00:00'
 '2016-06-05 00:00:00' '2016-06-08 00:00:00' '2016-06-07 00:00:00'
 '2016-06-10 00:00:00' '2016-06-13 00:00:00' '2016-06-16 00:00:00'
 '2016-06-18 00:00:00' '2016-06-17 00:00:00' '2016-06-19 00:00:00'
 '2016-06-21 00:00:00' '2016-06-22 00:00:00' '2016-06-23 00:00:00'
 '2016-06-24 00:00:00' '2016-06-29 00:00:00' '2016-07-02 00:00:00'
 '2016-07-04 00:00:00' '2016-07-08 00:00:00' '2016-07-07 00:00:00'
 '2016-07-09 00:00:00' '2016-07-10 00:00:00' '2016-07-11 00:00:00'
 '2016-07-14 00:00:00' '2016-07-15 00:00:00' '2016-07-16 00:00:00'
 '2016-07-18 00:00:00' '2016-07-20 00:00:00' '2016-07-21 00:00:00'
 '2016-07-23 00:00:00' '2016-07-27 00:00:00' '2016-07-29 00:00:00'
 '2016-07-30 00:00:00' '2016-08-02 00:00:00' '2016-08-01 00:00:00'
 '2016-08-04 00:00:00' '2016-08-11 00:00:00' '2016-08-12 00:00:00'
 '2016-08-14 00:00:00' '2016-08-15 00:00:00' '2016-08-18 00:00:00'
 '2016-08-19 00:00:00' '2016-08-22 00:00:00' '2016-08-24 00:00:00'
 '2016-08-25 00:00:00' '2016-08-29 00:00:00' '2016-08-27 00:00:00'
 '2016-08-30 00:00:00' '2016-09-01 00:00:00' '2016-09-02 00:00:00'
 '2016-09-04 00:00:00' '2016-09-03 00:00:00' '2016-09-06 00:00:00'
 '2016-09-05 00:00:00' '2016-09-13 00:00:00' '2016-09-12 00:00:00'
 '2016-09-15 00:00:00' '2016-09-17 00:00:00' '2016-09-16 00:00:00'
 '2016-09-20 00:00:00' '2016-09-21 00:00:00' '2016-09-22 00:00:00'
 '2016-09-27 00:00:00' '2016-09-29 00:00:00' '2016-09-30 00:00:00'
 '2016-10-01 00:00:00' '2016-10-03 00:00:00' '2016-10-04 00:00:00'
 '2016-10-08 00:00:00' '2016-10-10 00:00:00' '2016-10-11 00:00:00'
 '2016-10-13 00:00:00' '2016-10-18 00:00:00' '2016-10-20 00:00:00'
 '2016-10-23 00:00:00' '2016-10-24 00:00:00' '2016-10-26 00:00:00'
 '2016-10-27 00:00:00' '2016-10-29 00:00:00' '2016-11-04 00:00:00'
 '2016-11-08 00:00:00' '2016-11-11 00:00:00' '2016-11-13 00:00:00'
 '2016-11-19 00:00:00' '2016-11-21 00:00:00' '2016-11-23 00:00:00'
 '2016-11-25 00:00:00' '2016-11-28 00:00:00' '2016-11-29 00:00:00'
 '2016-11-30 00:00:00' '2016-12-01 00:00:00' '2016-12-08 00:00:00'
 '2016-12-09 00:00:00' '2016-12-10 00:00:00' '2016-12-12 00:00:00'
 '2016-12-13 00:00:00' '2016-12-15 00:00:00' '2016-12-16 00:00:00'
 '2016-12-19 00:00:00' '2016-12-23 00:00:00' '2016-12-22 00:00:00'
 '2016-12-26 00:00:00' '2016-12-28 00:00:00' '2016-12-30 00:00:00'
 '2016-12-31 00:00:00' '2017-01-02 00:00:00' '2017-01-05 00:00:00'
 '2017-01-06 00:00:00' '2017-01-07 00:00:00' '2017-01-08 00:00:00'
 '2017-01-09 00:00:00' '2017-01-10 00:00:00' '2017-01-12 00:00:00'
 '2017-01-14 00:00:00' '2017-01-17 00:00:00' '2017-01-20 00:00:00'
 '2017-01-21 00:00:00' '2017-01-23 00:00:00' '2017-01-24 00:00:00'
 '2017-01-25 00:00:00' '2017-01-27 00:00:00' '2017-01-29 00:00:00'
 '2017-01-28 00:00:00' '2017-01-31 00:00:00' '2017-02-01 00:00:00'
 '2017-02-04 00:00:00' '2017-02-05 00:00:00' '2017-02-07 00:00:00'
 '2017-02-08 00:00:00' '2017-02-09 00:00:00' '2017-02-13 00:00:00'
 '2017-02-14 00:00:00' '2017-02-15 00:00:00' '2017-02-16 00:00:00'
 '2017-02-17 00:00:00' '2017-02-23 00:00:00' '2017-02-25 00:00:00'
 '2017-02-26 00:00:00' '2017-02-27 00:00:00' '2017-03-01 00:00:00'
 '2017-03-02 00:00:00' '2017-03-04 00:00:00' '2017-03-06 00:00:00'
 '2017-03-08 00:00:00' '2017-03-09 00:00:00' '2017-03-10 00:00:00'
 '2017-03-15 00:00:00' '2017-03-18 00:00:00' '2017-03-22 00:00:00'
 '2017-03-25 00:00:00' '2017-03-31 00:00:00' '2017-04-04 00:00:00'
 '2017-04-05 00:00:00' '2017-04-07 00:00:00' '2017-04-06 00:00:00'
 '2017-04-10 00:00:00' '2017-04-08 00:00:00' '2017-04-11 00:00:00'
 '2017-04-13 00:00:00' '2017-04-12 00:00:00' '2017-04-23 00:00:00'
 '2017-04-19 00:00:00' '2017-04-25 00:00:00' '2017-04-24 00:00:00'
 '2017-04-28 00:00:00' '2017-04-29 00:00:00' '2017-04-30 00:00:00'
 '2017-05-05 00:00:00' '2017-05-06 00:00:00' '2017-05-10 00:00:00'
 '2017-05-16 00:00:00' '2017-05-17 00:00:00' '2017-05-18 00:00:00'
 '2017-05-19 00:00:00' '2017-05-23 00:00:00' '2017-05-30 00:00:00'
 '2017-06-04 00:00:00' '2017-06-09 00:00:00' '2017-06-11 00:00:00'
 '2017-06-14 00:00:00' '2017-06-15 00:00:00' '2017-06-17 00:00:00'
 '2017-06-18 00:00:00' '2017-06-24 00:00:00' '2017-06-20 00:00:00'
 '2017-06-23 00:00:00' '2017-06-19 00:00:00' '2017-06-22 00:00:00'
 '2017-06-29 00:00:00' '2017-07-04 00:00:00' '2017-07-05 00:00:00'
 '2017-07-06 00:00:00' '2017-07-09 00:00:00']

COUNTRIES

['Country_01' 'Country_02' 'Country_03']

LOCAL

['Local_01' 'Local_02' 'Local_03' 'Local_04' 'Local_05' 'Local_06'
 'Local_07' 'Local_08' 'Local_10' 'Local_09' 'Local_11' 'Local_12']

INDUSTRY SECTOR

['Mining' 'Metals' 'Others']

ACCIDENT LEVEL

['I' 'IV' 'III' 'II' 'V']

POTENTIAL ACCIDENT LEVEL

['IV' 'III' 'I' 'II' 'V' 'VI']

GENRE

['Male' 'Female']

EMPLOYEE OR THIRD PARTY

['Third Party' 'Employee' 'Third Party (Remote)']

CRITICAL RISK

['Pressed' 'Pressurized Systems' 'Manual Tools' 'Others'
 'Fall prevention (same level)' 'Chemical substances' 'Liquid Metal'
 'Electrical installation' 'Confined space'
 'Pressurized Systems / Chemical Substances'
 'Blocking and isolation of energies' 'Suspended Loads' 'Poll' 'Cut'
 'Fall' 'Bees' 'Fall prevention' '\nNot applicable' 'Traffic' 'Projection'
 'Venomous Animals' 'Plates' 'Projection/Burning' 'remains of choco'
 'Vehicles and Mobile Equipment' 'Projection/Choco' 'Machine Protection'
 'Power lock' 'Burn' 'Projection/Manual Tools'
 'Individual protection equipment' 'Electrical Shock'
 'Projection of fragments']

datadict = pd.DataFrame(data.dtypes)
datadict['MissingVal'] = data.isnull().sum()
datadict['NUnique']=data.nunique()
datadict

data.describe(include=['object'])

Country_01 is the country where most of the accidents happen (more than 50%)
Local_03 (which also belongs to Country_01) is where most of the accidents happen
Mining is also the most significant contributor to accidents
Male (95%) and Third Party (43%) also counts for kind of people that suffers more accident

industry_df = copy.deepcopy(data)

Rename columns to more Python standard naming¶

industry_df.drop("Unnamed: 0", axis=1, inplace=True)

industry_df.rename(columns={'Data':'date', 
                            'Countries':'country', 
                            'Local':'local',
                            'Industry Sector':'industry_sector',
                            'Accident Level':'accident_level',
                            'Potential Accident Level':'potential_accident_level',
                            'Genre':'gender', 
                            'Employee or Third Party':'emp_type',
                            'Critical Risk':'critical_risk',
                            'Description':'description'}
                   , inplace=True)


industry_df.head()

Check for duplicates¶

industry_df.duplicated().sum()

7

duplicates = industry_df.duplicated()

industry_df[duplicates]

industry_df.drop_duplicates(inplace=True)
industry_df.shape

(418, 10)

Preprocessing on 'Date' column¶

Split 'Date' to Year, Month, Day etc.

industry_df['date'] = pd.to_datetime(industry_df['date'])

industry_df['year'] = industry_df.date.apply(lambda x : x.year)
industry_df['month'] = industry_df.date.apply(lambda x : x.month)
industry_df['day'] = industry_df.date.apply(lambda x : x.day)
industry_df['weekday'] = industry_df.date.apply(lambda x : x.day_name())
industry_df['week_of_year'] = industry_df.date.apply(lambda x : x.weekofyear)

industry_df.head()

Add 'Season' column based on months

industry_df['season'] = industry_df['month'].apply(month2seasons)
industry_df.head(3)

Mark Offical Holidays for Brazil in the 'is_holiday' column as '1'

brazil_holidays = []

print(color.GREEN + color.BOLD + 'List of Brazil holidays in 2016' + color.END)
for date in holidays.Brazil(years = 2016).items():
    brazil_holidays.append(str(date[0]))
    print(date)

print(color.GREEN + color.BOLD + 'List of Brazil holidays in 2017' + color.END)
for date in holidays.Brazil(years = 2017).items():
    brazil_holidays.append(str(date[0]))
    print(date)

List of Brazil holidays in 2016
(datetime.date(2016, 1, 1), 'Ano novo')
(datetime.date(2016, 4, 21), 'Tiradentes')
(datetime.date(2016, 5, 1), 'Dia Mundial do Trabalho')
(datetime.date(2016, 9, 7), 'Independência do Brasil')
(datetime.date(2016, 10, 12), 'Nossa Senhora Aparecida')
(datetime.date(2016, 11, 2), 'Finados')
(datetime.date(2016, 11, 15), 'Proclamação da República')
(datetime.date(2016, 12, 25), 'Natal')
(datetime.date(2016, 3, 25), 'Sexta-feira Santa')
(datetime.date(2016, 3, 27), 'Páscoa')
(datetime.date(2016, 5, 26), 'Corpus Christi')
(datetime.date(2016, 2, 10), 'Quarta-feira de cinzas (Início da Quaresma)')
(datetime.date(2016, 2, 9), 'Carnaval')
List of Brazil holidays in 2017
(datetime.date(2017, 1, 1), 'Ano novo')
(datetime.date(2017, 4, 21), 'Tiradentes')
(datetime.date(2017, 5, 1), 'Dia Mundial do Trabalho')
(datetime.date(2017, 9, 7), 'Independência do Brasil')
(datetime.date(2017, 10, 12), 'Nossa Senhora Aparecida')
(datetime.date(2017, 11, 2), 'Finados')
(datetime.date(2017, 11, 15), 'Proclamação da República')
(datetime.date(2017, 12, 25), 'Natal')
(datetime.date(2017, 4, 14), 'Sexta-feira Santa')
(datetime.date(2017, 4, 16), 'Páscoa')
(datetime.date(2017, 6, 15), 'Corpus Christi')
(datetime.date(2017, 3, 1), 'Quarta-feira de cinzas (Início da Quaresma)')
(datetime.date(2017, 2, 28), 'Carnaval')

industry_df['is_holiday'] = [1 if str(val).split()[0] in brazil_holidays else 0 for val in industry_df['date']]
industry_df.head(3)

Exploratory Data Analysis¶

fig,ax = plt.subplots(nrows=1,ncols=2,figsize=(20,5))
acc_level = "Accident Level"
data[acc_level].reset_index().groupby(acc_level).count().sort_values(by= 
       "index").plot(kind="barh", legend=False, 
        ax=ax[0]).grid(axis='x')

pot_acc_level = "Potential Accident Level"
data[pot_acc_level].reset_index().groupby(pot_acc_level).count().sort_values(by= 
       "index").plot(kind="barh", legend=False, 
        ax=ax[1]).grid(axis='x')


plt.show()

fig = px.pie(industry_df, names='country', template='seaborn')
fig.update_traces(rotation=90, pull=[0.2,0.03,0.1,0.03,0.1], textinfo="percent+label", showlegend=False)
fig.show()

59% accidents reported in Country_01
31% accidents reported in Country_02
10% accidents reported in Country_03

fig = px.pie(industry_df, names='industry_sector', template='seaborn')
fig.update_traces(rotation=90, pull=[0.2,0.03,0.1,0.03,0.1], textinfo="percent+label", showlegend=False)
fig.show()

56.7% accidents reported in Mining
32.1% accidents reported in Metals
11.2% accidents reported in Others

fig = px.pie(industry_df, names='emp_type', template='seaborn')
fig.update_traces(rotation=90, pull=[0.2,0.03,0.1,0.03,0.1], textinfo="percent+label", showlegend=False)
fig.show()

44.3% accidents reported in Third_Party
42.6% accidents reported in Employee
13.2% accidents reported in Third Party (Remote)

plt.figure(figsize=(20,5))
descending_order = industry_df['critical_risk'].value_counts().sort_values(ascending=False).index
sns.countplot(x=industry_df['critical_risk'],order=descending_order)
plt.xticks(rotation = 'vertical')

(array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,
        17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32]),
 <a list of 33 Text xticklabel objects>)

Critical Risk 'Others' has been reported more that 200 times

fig = px.pie(data, names='Local', template='seaborn')
fig.update_traces(rotation=90, pull=[0.2,0.03,0.1,0.03,0.1], textinfo="percent+label", showlegend=False)
fig.show()

Local 3 has reported more that 20% of the total accidents

gender_cnt = np.round(industry_df['gender'].value_counts(normalize=True) * 100)

gender_cnt.plot(kind='bar', figsize=(5,4), width=0.9, cmap='gray', title='Gender')
#plt.bar(gender_cnt).opts(title="Gender Count", color="#8888ff", xlabel="Gender", ylabel="Percentage", yformatter='%d%%')\

<matplotlib.axes._subplots.AxesSubplot at 0x2b423e9f828>

year_cnt = np.round(industry_df['year'].value_counts(normalize=True,sort=False) * 100)
year_cnt.plot(kind='bar', figsize=(5,4), width=0.9, cmap='YlGn', title='Year count')

<matplotlib.axes._subplots.AxesSubplot at 0x2b423ee0278>

# collecting year in list to work with it.

y_r = industry_df['date'].dt.year
y_16 = len(y_r[y_r==2016])   # checking if there is 2016 in list and getting total numbers of it.
y_17 = len(y_r[y_r==2017])   # checking if there is 2017 in list and getting total numbers of it.
x_axis = ["2016", "2017"]
y_axis = [y_16,  y_17]
plt.bar(x_axis, y_axis, color = ['green', 'blue'])
plt.title("Accidents in 2016 and 2017")
plt.show()

The chart shows a declining trend in the number of accidents as 2016 and as 2017 .

month_cnt = np.round(industry_df['month'].value_counts(normalize=True,sort=False) * 100)
month_cnt.plot(kind='bar', figsize=(5,4), width=0.9, cmap='summer', title='Month Count')

<matplotlib.axes._subplots.AxesSubplot at 0x2b423f6c320>

This chart again shows a higher rate of accidents in the first 6 months of the year which can again be attributed to having twice the data as compared to the ast 6 months. However, it is important to note that month 2 (Feb) has a much higher count of accidents though the number of days (29) are less in that month.

day_cnt = np.round(industry_df['day'].value_counts(normalize=True,sort=False) * 100)
day_cnt.plot(kind='bar', figsize=(5,4), width=0.9, cmap='hsv', title='Day Count')

<matplotlib.axes._subplots.AxesSubplot at 0x2b424007be0>

Here, it is seen that the days - 4, 8 and 16 have the most number of accidents reported.

weekday_cnt = pd.DataFrame(np.round(industry_df['weekday'].value_counts(normalize=True,sort=False) * 100))
weekday_cnt['week_num'] = [['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday'].index(i) for i in weekday_cnt.index]
weekday_cnt.sort_values('week_num', inplace=True)

#hv.Bars((weekday_cnt.index, weekday_cnt.weekday)).opts(title="Weekday Count", color="#8888ff", xlabel="Weekdays") * hv.Curve(weekday_cnt['weekday']).opts(width=500, height=300, color='red', line_width=3

weekday_cnt['week_num'].plot(kind='bar', figsize=(5,4), width=0.9, cmap='pink', title='Weekday Count')

<matplotlib.axes._subplots.AxesSubplot at 0x2b423f2dba8>

Tuesdays and Thurdays have a higher than average accident count

# Helper function for relation between Accident Level/Potential Accident levels and other labels
def target_count(df, col1):
    fig = plt.figure(figsize = (15, 7.2))
    ax = fig.add_subplot(121)
    sns.countplot(x = col1, data = df, ax = ax, orient = 'v',
                  hue = 'accident_level').set_title(col1.capitalize() +' count plot by Accident Level', 
                                                                      fontsize = 13)
    plt.legend(labels = df['accident_level'].unique())
    plt.xticks(rotation = 90)
    
    ax = fig.add_subplot(122)
    sns.countplot(x = col1, data = df, ax = ax, orient = 'v', 
                  hue = 'potential_accident_level').set_title(col1.capitalize() +' count plot by Potential Accident Level', 
                                                                      fontsize = 13)
    plt.legend(labels = df['potential_accident_level'].unique())
    plt.xticks(rotation = 90)
    return plt.show()

target_count(industry_df, 'gender')

target_count(industry_df, 'emp_type')

target_count(industry_df, 'industry_sector')

target_count(industry_df, 'country')

target_count(industry_df, 'month')

target_count(industry_df, 'year')

target_count(industry_df, 'weekday')

The 'Accident Level' column is skewed with a lot of data for the less severe level 1. This can be seen in the huge disparity within the counts of the different level.
The 'Potential Accident Level' is more balance across all levels. Except for level VI which has only 1 record.

sns.countplot(x="country", data=industry_df, hue="emp_type")

<matplotlib.axes._subplots.AxesSubplot at 0x2b42412d2b0>

Country_01 has a lot of Third Party accidents followed by Employee accidents.
Country_02 has a lot of Employee accidents followed by Third Party (Remote) accidents.
Country_03 has no data for Third Party (Remote).

sns.countplot(x="country", data=industry_df, hue="industry_sector")

<matplotlib.axes._subplots.AxesSubplot at 0x2b42484e160>

Country_01 has a lot of Mining related accidents followed by Employee accidents.
Country_02 has a lot of Metal related accidents followed by Third Party (Remote) accidents.
Country_03 has only data for Others.

sns.countplot(x="country", data=industry_df,hue="gender")

<matplotlib.axes._subplots.AxesSubplot at 0x2b42456a518>

Country_03 has no data for Female

sns.countplot(x="emp_type", data=industry_df,hue="gender")

<matplotlib.axes._subplots.AxesSubplot at 0x2b4246626a0>

f = lambda x : np.round(x/x.sum() * 100)

ac_gen = industry_df.groupby(['gender','accident_level'])['accident_level'].count().unstack().apply(f, axis=1)
#ac = hv.Bars(pd.melt(ac_gen.reset_index(), ['gender']), ['gender','accident_level'], 'value').opts(opts.Bars(title="Accident Level by Gender Count"))

pot_ac_gen = industry_df.groupby(['gender','potential_accident_level'])['potential_accident_level'].count().unstack().apply(f, axis=1)
#pot_ac = hv.Bars(pd.melt(pot_ac_gen.reset_index(), ['gender']), ['gender','potential_accident_level'], 'value').opts(opts.Bars(title="Potential Accident Level by Gender Count"))

ac_gen.plot(kind='bar', figsize=(15,4), width=0.9, cmap='cool', title='Accident Level by Gender Count')
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)
pot_ac_gen.plot(kind='bar', figsize=(15,4), width=0.9, cmap='hot', title='Potential Accident Level by Gender Count')
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)




#(ac + pot_ac).opts(opts.Bars(width=400, height=300,tools=['hover'],show_grid=True,xrotation=0, ylabel="Percentage", yformatter='%d%%'))

<matplotlib.legend.Legend at 0x2b4259bde80>

f = lambda x : np.round(x/x.sum() * 100)

ac_em = industry_df.groupby(['emp_type','accident_level'])['accident_level'].count().unstack().apply(f, axis=1)
#ac = hv.Bars(pd.melt(ac_em.reset_index(), ['emp_type']), ['emp_type','accident_level'], 'value').opts(opts.Bars(title="Accident Level by Employee type Count"))

pot_ac_em = industry_df.groupby(['emp_type','potential_accident_level'])['potential_accident_level'].count().unstack().apply(f, axis=1)
#pot_ac = hv.Bars(pd.melt(pot_ac_em.reset_index(), ['emp_type']), ['emp_type','potential_accident_level'], 'value').opts(opts.Bars(title="Potential Accident Level by Employee type Count"))

ac_em.plot(kind='bar', figsize=(15,4), width=0.9, cmap='cool', title='Accident Level by Employee type Count')
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)
pot_ac_em.plot(kind='bar', figsize=(15,4), width=0.9, cmap='hot', title='Potential Accident Level by Employee type Count')
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)



#(ac + pot_ac).opts(opts.Bars(width=400, height=300,tools=['hover'],show_grid=True,xrotation=0, ylabel="Percentage", yformatter='%d%%',fontsize={'title':9}))

<matplotlib.legend.Legend at 0x2b425add358>

f = lambda x : np.round(x/x.sum() * 100)

ac_mo = industry_df.groupby(['month','accident_level'])['accident_level'].count().unstack().apply(f, axis=1).fillna(0)
#ac = hv.Curve(ac_mo['I'], label='I') * hv.Curve(ac_mo['II'], label='II') * hv.Curve(ac_mo['III'], label='III') * hv.Curve(ac_mo['IV'], label='IV') * hv.Curve(ac_mo['V'], label='V')\
        #.opts(opts.Curve(title="Accident Level by Month Count"))

pot_ac_mo = industry_df.groupby(['month','potential_accident_level'])['potential_accident_level'].count().unstack().apply(f, axis=1).fillna(0)
#pot_ac = hv.Curve(pot_ac_mo['I'], label='I') * hv.Curve(pot_ac_mo['II'], label='II') * hv.Curve(pot_ac_mo['III'], label='III') * hv.Curve(pot_ac_mo['IV'], label='IV')\
       # * hv.Curve(pot_ac_mo['V'], label='V') * hv.Curve(pot_ac_mo['VI'], label='VI').opts(opts.Curve(title="Potential Accident Level by Month Count"))



ac_mo.plot(kind='bar', figsize=(15,4), width=0.9, cmap='cool', title='Accident Level by Month Count')
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)
pot_ac_mo.plot(kind='bar', figsize=(15,4), width=0.9, cmap='hot', title='Potential Accident Level by Month Count')
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)


#hv.extension('bokeh')      
#(ac+pot_ac).opts(opts.Curve(width=800, height=300,tools=['hover'],show_grid=True, ylabel="Percentage", yformatter='%d%%')).cols(1)

<matplotlib.legend.Legend at 0x2b4263d0128>

f = lambda x : np.round(x/x.sum() * 100)

ac_weekday = industry_df.groupby(['weekday','accident_level'])['accident_level'].count().unstack().apply(f, axis=1).fillna(0)
ac_weekday['week_num'] = [['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday'].index(i) for i in ac_weekday.index]
ac_weekday.sort_values('week_num', inplace=True)
ac_weekday.drop('week_num', axis=1, inplace=True)
#ac = hv.Curve(ac_weekday['I'], label='I') * hv.Curve(ac_weekday['II'], label='II') * hv.Curve(ac_weekday['III'], label='III') * hv.Curve(ac_weekday['IV'], label='IV') * hv.Curve(ac_weekday['V'], label='V')\
        #.opts(opts.Curve(title="Accident Level by Weekday Count"))

pot_ac_weekday = industry_df.groupby(['weekday','potential_accident_level'])['potential_accident_level'].count().unstack().apply(f, axis=0).fillna(0)
pot_ac_weekday['week_num'] = [['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday'].index(i) for i in pot_ac_weekday.index]
pot_ac_weekday.sort_values('week_num', inplace=True)
pot_ac_weekday.drop('week_num', axis=1, inplace=True)
#pot_ac = hv.Curve(pot_ac_weekday['I'], label='I') * hv.Curve(pot_ac_weekday['II'], label='II') * hv.Curve(pot_ac_weekday['III'], label='III') * hv.Curve(pot_ac_weekday['IV'], label='IV')\
       # * hv.Curve(pot_ac_weekday['V'], label='V') * hv.Curve(pot_ac_weekday['VI'], label='VI').opts(opts.Curve(title="Potential Accident Level by Weekday Count"))

ac_weekday.plot(kind='bar', figsize=(15,4), width=0.9, cmap='cool', title='Accident Level by Weekday Count')
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)
pot_ac_weekday.plot(kind='bar', figsize=(15,4), width=0.9, cmap='hot', title='Potential Accident Level by Weekday Count')
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)

#hv.extension('bokeh')   
#(ac+pot_ac).opts(opts.Curve(width=800, height=300,tools=['hover'],show_grid=True, ylabel="Percentage", yformatter='%d%%')).cols(1)

<matplotlib.legend.Legend at 0x2b4265b1b38>

f = lambda x : np.round(x/x.sum() * 100)
ac_season = industry_df.groupby(['season','accident_level'])['accident_level'].count().unstack().apply(f, axis=1).fillna(0)
ac_season['season_num'] = [['Spring', 'Summer', 'Autumn', 'Winter'].index(i) for i in ac_season.index]
ac_season.sort_values('season_num', inplace=True)
ac_season.drop('season_num', axis=1, inplace=True)
#ac = hv.Curve(ac_season['I'], label='I') * hv.Curve(ac_season['II'], label='II') * hv.Curve(ac_season['III'], label='III') * hv.Curve(ac_season['IV'], label='IV') * hv.Curve(ac_season['V'], label='V')\
        #.opts(opts.Curve(title="Accident Level by Season Count"))

pot_ac_season = industry_df.groupby(['season','potential_accident_level'])['potential_accident_level'].count().unstack().apply(f, axis=0).fillna(0)
pot_ac_season['season_num'] = [['Spring', 'Summer', 'Autumn', 'Winter'].index(i) for i in pot_ac_season.index]
pot_ac_season.sort_values('season_num', inplace=True)
pot_ac_season.drop('season_num', axis=1, inplace=True)
#pot_ac = hv.Curve(pot_ac_season['I'], label='I') * hv.Curve(pot_ac_season['II'], label='II') * hv.Curve(pot_ac_season['III'], label='III') * hv.Curve(pot_ac_season['IV'], label='IV')\
        #* hv.Curve(pot_ac_season['V'], label='V') * hv.Curve(pot_ac_season['VI'], label='VI').opts(opts.Curve(title="Potential Accident Level by Season Count"))

ac_season.plot(kind='bar', figsize=(15,4), width=0.9, cmap='cool', title='Accident Level by Season Count')
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)
pot_ac_season.plot(kind='bar', figsize=(15,4), width=0.9, cmap='hot', title='Potential Accident Level by Season Count')
plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)

#hv.extension('bokeh')
#(ac+pot_ac).opts(opts.Curve(width=800, height=300,tools=['hover'],show_grid=True, ylabel="Percentage", yformatter='%d%%')).cols(1)

<matplotlib.legend.Legend at 0x2b426da94a8>

indexes = list(industry_df.loc[industry_df['description'].str.len() > 100, 'description'].index)
_ = industry_df.loc[indexes, 'accident_level'].value_counts().plot(kind = 'pie', autopct = '%.0f%%', labels = ['I', 'II', 'III', 'IV', 'V'], figsize = (10, 6))

_ = industry_df.loc[indexes, 'potential_accident_level'].value_counts().plot(kind = 'pie', autopct = '%.0f%%', labels = ['IV', 'III', 'II', 'I', 'V', 'VI'], figsize = (10, 6))

industry_df['description'].str.len().hist()

<matplotlib.axes._subplots.AxesSubplot at 0x2b426d07748>

industry_df['description'].str.split().map(lambda x: len(x)).hist()

<matplotlib.axes._subplots.AxesSubplot at 0x2b42726d978>

industry_df['description'].str.split().apply(lambda x : [len(i) for i in x]). \
   map(lambda x: np.mean(x)).hist()

<matplotlib.axes._subplots.AxesSubplot at 0x2b4277336d8>

nltk.download("stopwords")

[nltk_data] Downloading package stopwords to
[nltk_data]     C:\Users\G621315\AppData\Roaming\nltk_data...
[nltk_data]   Package stopwords is already up-to-date!

True

plot_top_stopwords_barchart(industry_df['description'])

def plot_top_non_stopwords_barchart(text):
    stop=set(stopwords.words('english'))
    
    new= text.str.split()
    new=new.values.tolist()
    corpus=[word for i in new for word in i]

    counter=Counter(corpus)
    most=counter.most_common()
    x, y=[], []
    for word,count in most[:40]:
        if (word not in stop):
            x.append(word)
            y.append(count)
            
    sns.barplot(x=y,y=x)

plot_top_non_stopwords_barchart(industry_df['description'])

def plot_top_ngrams_barchart(text, n=2):
    stop=set(stopwords.words('english'))

    new= text.str.split()
    new=new.values.tolist()
    corpus=[word for i in new for word in i]

    def _get_top_ngram(corpus, n=None):
        vec = CountVectorizer(ngram_range=(n, n)).fit(corpus)
        bag_of_words = vec.transform(corpus)
        sum_words = bag_of_words.sum(axis=0) 
        words_freq = [(word, sum_words[0, idx]) 
                      for word, idx in vec.vocabulary_.items()]
        words_freq =sorted(words_freq, key = lambda x: x[1], reverse=True)
        return words_freq[:10]

    top_n_bigrams=_get_top_ngram(text,n)[:10]
    x,y=map(list,zip(*top_n_bigrams))
    sns.barplot(x=y,y=x)
    
plot_top_ngrams_barchart(industry_df['description'],2)

plot_top_ngrams_barchart(industry_df['description'],3)

plot_top_ngrams_barchart(industry_df['description'],4)

Preprocessing for Description column¶

industry_df = get_cleaned_desc(industry_df, 'description', 'cleaned_description')

Converting description to lower case
Replacing apostrophes to the standard lexicons
Removing punctuations
Applying Lemmatizer
Removing multiple spaces between words
Removing stop words

wordcloud = WordCloud(width = 1500, height = 800, random_state=0, background_color='black', colormap='rainbow', \
                      min_font_size=5, max_words=300, collocations=False).generate(" ".join(industry_df['cleaned_description'].values))
plt.figure(figsize=(15,10))
plt.imshow(wordcloud)
plt.axis('off')
plt.show()

# Get length of each line
industry_df['line_length'] = industry_df['cleaned_description'].str.len()

print(color.GREEN + color.BOLD + 'Minimum line length: ' + color.END + str(industry_df['line_length'].min()))
print(color.GREEN + color.BOLD + 'Maximum line length: ' + color.END + str(industry_df['line_length'].max()))
print(color.GREEN + color.BOLD + 'Line with maximum length: ' + color.END + str(industry_df[industry_df['line_length'] == industry_df['line_length'].max()]['cleaned_description'].values[0]))


# Get length of each word
industry_df['nb_words'] = industry_df['cleaned_description'].apply(lambda x: len(x.split(' ')))

print(color.BLUE + color.BOLD + 'Minimum number of words:  ' + color.END + str(industry_df['nb_words'].min()))
print(color.BLUE + color.BOLD + 'Maximum number of words:  ' + color.END + str(industry_df['nb_words'].max()))
print(color.BLUE + color.BOLD + 'Line with maximum number of words:  ' + color.END + str(industry_df[industry_df['nb_words'] == industry_df['nb_words'].max()]['cleaned_description'].values[0]))

Minimum line length: 64
Maximum line length: 680
Line with maximum length: level 3490 gallery 239 holding activity bolter equipment 24 operator performs drilling first hole support right gable 7footdeep drill end drill rod breaks leaving threads inside drilling machines shank operator assistant decide make two empty percussion attempts free threads shank without success third attempt assistant enters 38 corrugated iron central hole rest bar embedded shank generate pressure moment operator activates percussion generates movement shank hits palm victims left hand generating described injury worker wearing safety gloves time accident end corrugated iron contact left hand shaped like cane worker time accident positioned roof supported mesh split set
Minimum number of words:  10
Maximum number of words:  98
Line with maximum number of words:  performing sleeve removal maneuver hole 600 meters deep general silva pressed one side locking nut rods together jack holds entire weight rods maneuver locking procedure not effective weight rods secured steel wire rope probe winch moment driller pedro released brake winch inefficacy locking done one side chestnut without aid monkey caused sliding rods auxiliary prepared manual unlocking rods holding faucet key firmly probe tower composition shifted stems slid hands shifted downward causing left hand strike base probe tower structure causing cuts 4th 5th quirodactyl employee taken hospital went medical care wound sutured 16 stitches removed 10 days activities

print(color.GREEN + 'Five point summary for number of words' + color.END)
display(industry_df['nb_words'].describe().round(0).astype(int)); 

print(color.GREEN+ '99% quantilie: ' + color.END + str(industry_df['nb_words'].quantile(0.99)))

Five point summary for number of words

count    418
mean      34
std       16
min       10
25%       21
50%       31
75%       43
max       98
Name: nb_words, dtype: int32

99% quantilie: 83.65999999999997

# define training data
sentences = industry_df['cleaned_description']

# train model
model = Word2Vec(sentences, min_count=1)

# summarize the loaded model
print(color.BLUE+ "Word2Vec Model: " + color.END + str(model))

# summarize vocabulary
words = list(model.wv.index_to_key)
print(color.BLUE+ "Words: " + color.END + str(words))

# save model
model.save('model.bin')
print(color.GREEN+ "Model Saved successfully" + color.END)

# load model
new_model = Word2Vec.load('model.bin')
print(color.BLUE+ "Model loaded from Disk: " + color.END + str(new_model))

Word2Vec Model: Word2Vec(vocab=48, vector_size=100, alpha=0.025)
Words: [' ', 'e', 'i', 't', 'r', 'a', 'n', 'o', 's', 'l', 'c', 'd', 'p', 'm', 'g', 'u', 'h', 'f', 'b', 'y', 'v', 'k', 'w', '0', '1', 'x', 'j', '2', '5', '4', '3', 'q', '8', '7', '6', 'z', '9', '\u200b', '°', 'á', 'é', 'ã', 'ó', 'â', 'í', '¼', 'ô', 'ú']
Model Saved successfully
Model loaded from Disk: Word2Vec(vocab=48, vector_size=100, alpha=0.025)

embeddings_index = {}
EMBEDDING_FILE = 'glove.6B.200d.txt'
f = open(EMBEDDING_FILE, encoding="utf8")
for line in tqdm(f):
    values = line.split()
    word = values[0]
    coefs = np.asarray(values[1:], dtype='float32')
    embeddings_index[word] = coefs
f.close()

print(color.GREEN + '\nWord vectors Found:' + color.END + str(len(embeddings_index)))

400000it [00:50, 7992.17it/s]


Word vectors Found:400000

# this function creates a normalized vector for the whole sentence
def sent2vec(s):
    words = str(s).lower()
    words = word_tokenize(words)
    words = [w for w in words if not w in stop_words]
    words = [w for w in words if w.isalpha()]
    M = []
    for w in words:
        try:
            M.append(embeddings_index[w])
        except:
            continue
    M = np.array(M)
    v = M.sum(axis=0)
    if type(v) != np.ndarray:
        return np.zeros(300)
    return v / np.sqrt((v ** 2).sum())

# create sentence GLOVE embeddings vectors using the above function for training and validation set
industry_glove_df = [sent2vec(x) for x in tqdm(industry_df['cleaned_description'])]

100%|██████████████████████████████████████████████████████████████████████████████| 418/418 [00:00<00:00, 1151.22it/s]

industry_glove_df[0]

array([ 1.39856245e-03,  8.99862126e-02,  1.08799164e-03, -3.09016760e-02,
        6.19197078e-03,  6.87033385e-02, -1.02347635e-01, -2.87667587e-02,
       -6.25691786e-02,  4.17288505e-02, -1.61288977e-02,  1.33184874e-02,
        1.12336598e-01,  1.67398853e-03,  8.71078447e-02, -1.93325020e-02,
       -3.66557874e-02, -2.86911987e-02,  7.09104389e-02,  8.07439629e-03,
       -4.30122279e-02,  6.14812076e-01,  9.06249657e-02,  1.87851384e-03,
        3.43448929e-02,  3.48158218e-02, -3.08507830e-02,  2.58244704e-02,
       -3.18089649e-02, -6.93444610e-02,  5.13918102e-02, -4.53802906e-02,
        1.09138573e-02, -1.77216530e-02,  1.88153088e-02, -3.13805038e-04,
       -8.93176496e-02, -6.43905252e-02, -7.05620348e-02,  4.99016382e-02,
        6.68967739e-02,  3.37204821e-02, -1.01575870e-02,  6.47136495e-02,
        3.32460664e-02,  6.59773797e-02,  4.91438136e-02,  2.39142426e-03,
        4.29478846e-02,  7.00706914e-02,  3.60498391e-02,  1.07721733e-02,
       -5.63761368e-02, -3.21689770e-02,  7.09457174e-02, -3.54300961e-02,
       -3.92932072e-02, -3.85472141e-02, -2.62974612e-02,  4.47593592e-02,
        1.50641687e-02, -5.77454120e-02, -4.51208316e-02, -3.04190014e-02,
        9.40906536e-03, -5.03021665e-03, -5.06151430e-02,  1.01581053e-03,
        2.60779001e-02, -1.67292245e-02,  1.17642313e-01,  2.31643505e-02,
        4.17141542e-02, -1.84773505e-02, -9.88877714e-02,  6.35349378e-02,
       -1.37783531e-02, -5.16787032e-03,  1.22039374e-02,  9.61354375e-03,
        3.04872207e-02,  2.13078726e-02,  7.58600375e-03,  4.15864587e-03,
        3.28518599e-02, -1.07852377e-01, -3.05358749e-02, -1.67166851e-02,
        1.49722114e-01, -1.88600883e-01,  4.67074588e-02, -2.90540047e-02,
        1.68920830e-02,  7.29843080e-02,  1.67668313e-02,  3.49243172e-02,
        6.41853064e-02, -1.02650691e-02, -4.03477475e-02, -2.05384288e-02,
        4.32290277e-03, -2.80256215e-02, -1.58334970e-02,  1.84803884e-02,
        7.06879701e-03, -2.71190237e-02,  1.99222136e-02,  1.83336079e-01,
        1.23815555e-02, -3.45078893e-02, -2.92782132e-02, -4.45271134e-02,
        1.34143839e-03,  5.90255344e-03, -2.51722196e-03, -2.13238988e-02,
        5.28019406e-02, -2.98531409e-02, -5.86500429e-02,  2.01110952e-02,
        9.56291258e-02, -4.88960184e-02,  5.44730108e-03,  5.20819947e-02,
       -4.53534164e-02, -3.48782912e-02,  1.03540562e-01, -1.08262217e-02,
       -3.61204445e-02,  3.25142406e-02, -3.47124413e-02,  1.45224586e-01,
        2.03661472e-02, -1.04528792e-01, -2.18481813e-02,  4.74526621e-02,
       -9.77941323e-03, -4.29209089e-03,  1.69468895e-02, -3.78255025e-02,
        2.16276869e-02,  4.54882346e-02,  3.56075428e-02, -5.51389828e-02,
        2.60926038e-01,  1.82563942e-02, -4.47200984e-02, -2.86088791e-02,
       -1.59378573e-02, -5.30578382e-03, -1.09551307e-02,  2.62108631e-02,
        3.54089551e-02, -1.17622353e-02, -4.30766270e-02,  1.52955484e-02,
       -5.43822683e-02,  7.38786831e-02,  2.70258598e-02, -1.80050641e-01,
        5.00808023e-02, -2.57551540e-02, -3.60159948e-02,  3.79033387e-03,
       -1.88139044e-02, -2.77041830e-02, -4.53292727e-02,  3.66431358e-03,
       -2.66571213e-02,  9.38335359e-02,  3.85193042e-02, -1.25983190e-02,
        1.44656068e-02,  2.56294347e-02,  3.38767975e-04, -4.31163609e-02,
       -2.62325089e-02, -7.10816309e-02,  2.32091248e-02, -1.49043193e-02,
        2.15159059e-01, -6.13922514e-02, -3.65052884e-03,  2.37665121e-02,
        1.44381048e-02, -3.26929837e-02, -5.17858565e-02, -9.15953703e-03,
        2.62024682e-02,  8.20122361e-02, -5.05710766e-02,  4.56798170e-03,
       -3.44055928e-02,  4.85826991e-02,  1.09664332e-02,  1.60338590e-03,
        2.03831252e-02, -1.53971147e-02, -5.91683984e-02,  5.48761599e-02],
      dtype=float32)

Get top 30 Terms as columns

industry_tfidf_df = pd.DataFrame()
for i in [1,2,3]:
    vec_tfidf = TfidfVectorizer(max_features=10, norm='l2', stop_words='english', lowercase=True, use_idf=True, ngram_range=(i,i))
    X = vec_tfidf.fit_transform(industry_df['cleaned_description']).toarray()
    tfs = pd.DataFrame(X, columns=["tfidf_" + n for n in vec_tfidf.get_feature_names()])
    industry_tfidf_df = pd.concat([industry_tfidf_df.reset_index(drop=True), tfs.reset_index(drop=True)], axis=1)

industry_tfidf_df.head(3)

# To replace white space everywhere in Employee type
industry_df['emp_type'] = industry_df['emp_type'].str.replace(' ', '_')
industry_df['emp_type'].value_counts()

Third_Party             185
Employee                178
Third_Party_(Remote)     55
Name: emp_type, dtype: int64

# To replace white space everywhere in Critical Risk
industry_df['critical_risk'] = industry_df['critical_risk'].str.replace('\n', '').str.replace(' ', '_')
industry_df['critical_risk'].value_counts().head()

Others                 229
Pressed                 24
Manual_Tools            20
Chemical_substances     17
Cut                     14
Name: critical_risk, dtype: int64

Label Encoding for Categorical variables¶

# Create Industry DataFrame
industry_featenc_df = industry_df[['year','month','day','week_of_year']].reset_index(drop=True)

# Label encoding
industry_df['season'] = industry_df['season'].replace('Summer', 'aSummer').replace('Autumn', 'bAutumn').replace('Winter', 'cWinter').replace('Spring', 'dSpring')
industry_featenc_df['season'] = LabelEncoder().fit_transform(industry_df['season']).astype(np.int8)

industry_df['weekday'] = industry_df['weekday'].replace('Monday', 'aMonday').replace('Tuesday', 'bTuesday').replace('Wednesday', 'cWednesday').replace('Thursday', 'dThursday').replace('Friday', 'eFriday').replace('Saturday', 'fSaturday').replace('Sunday', 'gSunday')
industry_featenc_df['weekday'] = LabelEncoder().fit_transform(industry_df['weekday']).astype(np.int8)

industry_featenc_df['country'] = LabelEncoder().fit_transform(industry_df['country']).astype(np.int8)
industry_featenc_df['local'] = LabelEncoder().fit_transform(industry_df['local']).astype(np.int8)
industry_featenc_df['gender'] = LabelEncoder().fit_transform(industry_df['gender']).astype(np.int8)
industry_featenc_df['industry_sector'] = LabelEncoder().fit_transform(industry_df['industry_sector']).astype(np.int8)
industry_featenc_df['emp_type'] = LabelEncoder().fit_transform(industry_df['emp_type']).astype(np.int8)
industry_featenc_df['critical_risk'] = LabelEncoder().fit_transform(industry_df['critical_risk']).astype(np.int8)

industry_featenc_df['accident_level'] = LabelEncoder().fit_transform(industry_df['accident_level']).astype(np.int8)
industry_featenc_df['potential_accident_level'] = LabelEncoder().fit_transform(industry_df['potential_accident_level'].replace('I', 'aI').replace('II', 'bII').replace('III', 'cIII').replace('IV', 'dIV').replace('V', 'eV').replace('VI', 'fVI')).astype(np.int8)
industry_featenc_df.head(3)

industry_df['accident_level'].unique()

array(['I', 'IV', 'III', 'II', 'V'], dtype=object)

# convert integers to dummy variables (i.e. one hot encoded)
dummy_y = np_utils.to_categorical(industry_featenc_df['accident_level'])
dummy_y

array([[1., 0., 0., 0., 0.],
       [1., 0., 0., 0., 0.],
       [1., 0., 0., 0., 0.],
       ...,
       [1., 0., 0., 0., 0.],
       [1., 0., 0., 0., 0.],
       [1., 0., 0., 0., 0.]], dtype=float32)

np.any(np.isnan(industry_featenc_df))

False

industry_feat_df = industry_featenc_df.join(pd.DataFrame(industry_glove_df).iloc[:,0:30].reset_index(drop=True))

industry_feat_df = industry_featenc_df.join(industry_tfidf_df.reset_index(drop=True))

industry_feat_df.head()

Display old accident level counts

industry_feat_df['accident_level'].value_counts()

0    309
1     40
2     31
3     30
4      8
Name: accident_level, dtype: int64

Test-Train split¶

X = industry_feat_df.drop(['accident_level','potential_accident_level'], axis = 1) # Considering all Predictors
y = industry_feat_df['accident_level']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.20, random_state = 1, stratify = y)
X_train, X_test, y_train_dummy, y_test_dummy = train_test_split(X, dummy_y, test_size = 0.20, random_state = 1, stratify = y)

print(color.BLUE + 'X_train shape: ' + color.END + str((X_train.shape[0], X_train.shape[1])))
print(color.BLUE + 'y_train shape: ' + color.END + str(y_train.shape[0]))
print(color.BLUE + 'X_test shape: ' + color.END + str((X_test.shape[0], X_test.shape[1])))
print(color.BLUE + 'y_test shape: ' + color.END + str(y_test.shape[0]))

X_train shape: (334, 42)
y_train shape: 334
X_test shape: (84, 42)
y_test shape: 84

Upsampling¶

# Concatenate our training data back together
X_up = pd.concat([X_train, y_train], axis=1)

# Get the majority and minority class
pot_acc_level_0_majority = X_up[X_up['accident_level'] == 0]
pot_acc_level_1_minority = X_up[X_up['accident_level'] == 1]
pot_acc_level_2_minority = X_up[X_up['accident_level'] == 2]
pot_acc_level_3_minority = X_up[X_up['accident_level'] == 3]
pot_acc_level_4_minority = X_up[X_up['accident_level'] == 4]


# Upsample Level2 minority class
pot_acc_level_1_upsampled = resample(pot_acc_level_1_minority,
                                replace = True, # sample with replacement
                                n_samples = len(pot_acc_level_0_majority), # to match majority class
                                random_state = 1)

# Upsample Level3 minority class
pot_acc_level_2_upsampled = resample(pot_acc_level_2_minority,
                                replace = True, # sample with replacement
                                n_samples = len(pot_acc_level_0_majority), # to match majority class
                                random_state = 1)

# Upsample Level4 minority class
pot_acc_level_3_upsampled = resample(pot_acc_level_3_minority,
                                replace = True, # sample with replacement
                                n_samples = len(pot_acc_level_0_majority), # to match majority class
                                random_state = 1)


# Upsample Level5 minority class
pot_acc_level_4_upsampled = resample(pot_acc_level_4_minority,
                                replace = True, # sample with replacement
                                n_samples = len(pot_acc_level_0_majority), # to match majority class
                                random_state = 1)


industry_df_upsampled = pd.concat([pot_acc_level_0_majority, pot_acc_level_1_upsampled, pot_acc_level_2_upsampled, pot_acc_level_3_upsampled, pot_acc_level_4_upsampled])

industry_df_upsampled['accident_level'].value_counts()

4    247
3    247
2    247
1    247
0    247
Name: accident_level, dtype: int64

# Separate input features and target
X_train_up = industry_df_upsampled.drop(['accident_level'], axis = 1) # Considering all Predictors
y_train_up = industry_df_upsampled['accident_level']

Smote¶

sm = SMOTE(random_state=1)
X_train_smote, y_train_smote = sm.fit_resample(X_train, y_train)
df_smote = pd.concat([pd.DataFrame(X_train_smote), pd.DataFrame(y_train_smote)], axis=1)

df_smote.head()

columns_list=[]
for col in industry_df_upsampled.columns:
    columns_list.append(col)

df_smote.columns = columns_list

df_smote.head()

# Separate input features and target
X_train_smote = df_smote.iloc[:,:-1] # Considering all Predictors
y_train_smote = df_smote.iloc[:,-1:]
X_train_smote.head(1)

y_train_smote['accident_level'].value_counts()

4    247
3    247
2    247
1    247
0    247
Name: accident_level, dtype: int64

# convert integers to dummy variables (i.e. one hot encoded)
y_train_smote_dummy = np_utils.to_categorical(y_train_smote['accident_level'])
y_train_smote_dummy

array([[0., 1., 0., 0., 0.],
       [0., 1., 0., 0., 0.],
       [1., 0., 0., 0., 0.],
       ...,
       [0., 0., 0., 0., 1.],
       [0., 0., 0., 0., 1.],
       [0., 0., 0., 0., 1.]], dtype=float32)

Scaling¶

# Transform independent features
scaler_X = StandardScaler()#StandardScaler()
pipeline = Pipeline(steps=[('s', scaler_X)])
X_train.iloc[:,:12] = pipeline.fit_transform(X_train.iloc[:,:12]) # Scaling only first 12 feautres

X_test.iloc[:,:12] = pipeline.fit_transform(X_test.iloc[:,:12]) # Scaling only first 12 feautres

X_train_up.iloc[:,:12] = pipeline.fit_transform(X_train_up.iloc[:,:12]) # Scaling only first 12 feautres
X_train_smote.iloc[:,:12] = pipeline.fit_transform(X_train_smote.iloc[:,:12]) # Scaling only first 12 feautres

X_train.head(3)

# Transform independent features
scaler_X = StandardScaler()#StandardScaler()
pipeline = Pipeline(steps=[('s', scaler_X)])

Principal Component Analysis¶

# generating the covariance matrix and the eigen values for the PCA analysis
cov_matrix = np.cov(X_train.T) # the relevanat covariance matrix
print(color.GREEN + 'Covariance Matrix: \n' + color.END + str(cov_matrix))

#generating the eigen values and the eigen vectors
e_vals, e_vecs = np.linalg.eig(cov_matrix)
print(color.GREEN + '\nEigenvectors: \n' + color.END + str(e_vecs))
print(color.GREEN + '\nEigenvalues \n' + color.END + str(e_vals))

Covariance Matrix: 
[[ 1.00300300e+00 -4.24727770e-01  2.53271550e-02 ... -2.06945085e-03
  -4.89406968e-03 -5.68968778e-03]
 [-4.24727770e-01  1.00300300e+00  2.68776694e-02 ...  7.08759577e-03
   4.65172198e-03  5.42953855e-03]
 [ 2.53271550e-02  2.68776694e-02  1.00300300e+00 ... -1.28805019e-03
  -5.08213127e-03 -3.28488592e-03]
 ...
 [-2.06945085e-03  7.08759577e-03 -1.28805019e-03 ...  6.86698871e-03
  -7.05215662e-05 -1.74059396e-04]
 [-4.89406968e-03  4.65172198e-03 -5.08213127e-03 ... -7.05215662e-05
   5.61025313e-03 -1.20024227e-04]
 [-5.68968778e-03  5.42953855e-03 -3.28488592e-03 ... -1.74059396e-04
  -1.20024227e-04  1.63635526e-02]]

Eigenvectors: 
[[-3.50985509e-01  1.07572835e-01 -1.86205860e-01 ... -7.31043981e-03
   3.17588199e-03  6.73034750e-04]
 [ 5.60434833e-01 -2.76065216e-02 -6.26074591e-02 ... -2.36707634e-02
  -2.37858320e-02 -1.34427231e-03]
 [ 1.95724611e-02  1.26937026e-02 -4.77280078e-01 ...  4.67886712e-03
  -5.90609515e-04  9.54800459e-03]
 ...
 [ 3.12576650e-03 -4.15770297e-03  2.38677909e-03 ...  2.73712070e-03
   1.82243590e-02 -1.83915235e-02]
 [ 3.39378227e-03 -1.84947664e-03  3.37660650e-04 ... -7.69340795e-02
   9.57429221e-02  1.28760588e-02]
 [ 3.69706058e-03 -3.46086866e-03 -3.52159508e-03 ... -1.28345079e-01
  -5.13498286e-01  5.09599351e-01]]

Eigenvalues 
[2.85335836e+00 1.93574339e+00 1.37641102e+00 1.09440302e+00
 1.02760376e+00 9.45877581e-01 8.76178081e-01 7.41806390e-01
 6.21729966e-01 4.22030498e-01 2.69203500e-01 1.44132929e-01
 1.28703323e-01 1.19339936e-01 9.85243402e-02 6.99159533e-02
 6.70693432e-02 5.93131509e-02 5.84102947e-02 4.88931769e-02
 4.39961683e-02 4.01709475e-02 3.86969563e-02 3.56757977e-02
 3.06365961e-02 3.01858670e-02 2.65814918e-02 1.47404007e-03
 1.94623424e-03 3.22234748e-03 2.33838336e-02 5.17279197e-03
 7.05411094e-03 8.13692751e-03 8.25450891e-03 1.01957300e-02
 2.05421427e-02 1.92682368e-02 1.32878402e-02 1.64659615e-02
 1.54800084e-02 1.44933454e-02]

# the "cumulative variance explained" analysis 
tot = sum(e_vals)
var_exp = [( i /tot ) * 100 for i in sorted(e_vals, reverse=True)]
cum_var_exp = np.cumsum(var_exp)
print(color.GREEN + "Cumulative Variance Explained" + color.END + str(cum_var_exp))

Cumulative Variance Explained[ 21.336759    35.81180385  46.1042896   54.28798424  61.97216933
  69.04522482  75.59708343  81.14414139  85.79329539  88.94914261
  90.96218461  92.03997758  93.00239149  93.89478815  94.63153055
  95.15434601  95.65587516  96.09940529  96.53618407  96.90179603
  97.23078928  97.53117841  97.82054536  98.08732079  98.31641423
  98.54213721  98.74090751  98.91576645  99.06937588  99.21345933
  99.33658801  99.45234396  99.56072188  99.6600853   99.73632662
  99.79805195  99.85889802  99.91164705  99.950328    99.97442397
  99.98897747 100.        ]

# Plotting the variance expalained by the principal components and the cumulative variance explained.
plt.figure(figsize=(20 , 5))
plt.bar(range(1, e_vals.size + 1), var_exp, alpha = 0.5, align = 'center', label = 'Individual explained variance')
plt.step(range(1, e_vals.size + 1), cum_var_exp, where='mid', label = 'Cumulative explained variance')
plt.ylabel('Explained Variance Ratio')
plt.xlabel('Principal Components')
plt.legend(loc = 'best')
plt.tight_layout()
plt.show()

Capturing 90% variance of the data

pca = PCA(n_components = 0.90)
X_train_reduced = pca.fit_transform(X_train)
X_test_reduced = pca.transform(X_test)

print(X_train_reduced.shape)
print(X_test_reduced.shape)

(334, 11)
(84, 11)

Design, train and test Machine Learning classifiers¶

Set benchmark¶

# DummyClassifier to predict all Accident levels
dummy = DummyClassifier(strategy='stratified').fit(X_train, y_train)
dummy_pred = dummy.predict(X_test)

# checking unique labels
print(color.GREEN + 'Unique predicted labels: ' + color.END + str(np.unique(dummy_pred)))

# checking accuracy
print(color.GREEN + 'Test score: ' + color.END + str(accuracy_score(y_test, dummy_pred)))

Unique predicted labels: [0 1 2 3 4]
Test score: 0.5119047619047619

yhat_probs = dummy.predict(X_test)
yhat_classes = (np.asarray(dummy.predict(X_test))).round()
accuracy = accuracy_score(y_test, yhat_classes)
precision = precision_score(y_test, yhat_classes, average='micro')
recall = recall_score(y_test, yhat_classes, average='micro')
f1 = f1_score(y_test, yhat_classes, average='micro')
print(color.BLUE + 'Accuracy: ' + color.END + str(accuracy))
print(color.BLUE + 'Precision: ' + color.END + str(precision))
print(color.BLUE + 'Recall: ' + color.END +  str(recall))
print(color.BLUE + 'F1 score: ' + color.END + str(f1))

Accuracy: 0.5952380952380952
Precision: 0.5952380952380952
Recall: 0.5952380952380952
F1 score: 0.5952380952380952

Data with Label Encoding¶

train_test_allmodels(X_train, X_test, y_train, y_test, 'no')

[22:24:54] WARNING: C:/Users/Administrator/workspace/xgboost-win64_release_1.3.0/src/learner.cc:1061: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.

Data with Upsampling¶

train_test_allmodels(X_train_up, X_test, y_train_up, y_test, 'no')

[22:24:59] WARNING: C:/Users/Administrator/workspace/xgboost-win64_release_1.3.0/src/learner.cc:1061: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.

Data with SMOTE¶

train_test_allmodels(X_train_smote, X_test, y_train_smote, y_test, 'no')

[22:25:06] WARNING: C:/Users/Administrator/workspace/xgboost-win64_release_1.3.0/src/learner.cc:1061: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.

Design, train and test Neural networks classifiers¶

Numeric Classification - Data with Label Encoded Target Variable¶

# fix random seed for reproducibility
seed = 7
np.random.seed(seed)
tf.random.set_seed(seed)

# define the model
model = Sequential()
model.add(Dense(50, input_dim=X_train.shape[1], activation='relu', kernel_initializer='he_uniform'))
model.add(Dense(100, activation='relu', kernel_initializer='he_uniform'))
model.add(Dense(150, activation='relu', kernel_initializer='he_uniform'))
model.add(Dense(40, activation='relu', kernel_initializer='he_uniform'))
model.add(Dense(1, activation='linear'))

# compile the keras model
#opt = optimizers.Adam(lr=1e-3)
opt = SGD(lr=0.001, momentum=0.9)
model.compile(loss='mse', optimizer=opt, metrics=['accuracy'])

# Use earlystopping
callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=5, min_delta=0.001)
rlrp = ReduceLROnPlateau(monitor='val_loss', factor=0.0001, patience=5, min_delta=1E-4)

# fit the keras model on the dataset
training_history = model.fit(X_train, y_train, epochs=100, batch_size=8, verbose=1, validation_data=(X_test, y_test), callbacks=[rlrp])

Epoch 1/100
42/42 [==============================] - 5s 108ms/step - loss: 1.2392 - accuracy: 0.4675 - val_loss: 0.9094 - val_accuracy: 0.4643
Epoch 2/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8253 - accuracy: 0.4888 - val_loss: 0.9694 - val_accuracy: 0.6667
Epoch 3/100
42/42 [==============================] - 0s 6ms/step - loss: 1.0125 - accuracy: 0.5756 - val_loss: 0.9071 - val_accuracy: 0.5833
Epoch 4/100
42/42 [==============================] - 0s 6ms/step - loss: 0.8720 - accuracy: 0.4272 - val_loss: 0.9419 - val_accuracy: 0.4762
Epoch 5/100
42/42 [==============================] - 0s 6ms/step - loss: 0.6629 - accuracy: 0.5380 - val_loss: 0.9377 - val_accuracy: 0.5714
Epoch 6/100
42/42 [==============================] - 0s 8ms/step - loss: 0.6298 - accuracy: 0.5937 - val_loss: 0.9888 - val_accuracy: 0.4048
Epoch 7/100
42/42 [==============================] - 0s 7ms/step - loss: 0.5741 - accuracy: 0.6397 - val_loss: 0.9343 - val_accuracy: 0.4762
Epoch 8/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5052 - accuracy: 0.5787 - val_loss: 0.9967 - val_accuracy: 0.6071
Epoch 9/100
42/42 [==============================] - 0s 10ms/step - loss: 0.5071 - accuracy: 0.6836 - val_loss: 0.9898 - val_accuracy: 0.5000
Epoch 10/100
42/42 [==============================] - 1s 14ms/step - loss: 0.4776 - accuracy: 0.6008 - val_loss: 0.9907 - val_accuracy: 0.5000
Epoch 11/100
42/42 [==============================] - 0s 8ms/step - loss: 0.5381 - accuracy: 0.5831 - val_loss: 0.9906 - val_accuracy: 0.5000
Epoch 12/100
42/42 [==============================] - 0s 8ms/step - loss: 0.4933 - accuracy: 0.6093 - val_loss: 0.9906 - val_accuracy: 0.5000
Epoch 13/100
42/42 [==============================] - 0s 9ms/step - loss: 0.5316 - accuracy: 0.6121 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 14/100
42/42 [==============================] - 0s 7ms/step - loss: 0.4423 - accuracy: 0.6432 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 15/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5214 - accuracy: 0.5849 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 16/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4795 - accuracy: 0.6010 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 17/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5946 - accuracy: 0.5454 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 18/100
42/42 [==============================] - 0s 6ms/step - loss: 0.4750 - accuracy: 0.6178 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 19/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4138 - accuracy: 0.6353 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 20/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5617 - accuracy: 0.5907 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 21/100
42/42 [==============================] - ETA: 0s - loss: 0.5326 - accuracy: 0.61 - 0s 6ms/step - loss: 0.5261 - accuracy: 0.6078 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 22/100
42/42 [==============================] - 1s 13ms/step - loss: 0.5802 - accuracy: 0.5533 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 23/100
42/42 [==============================] - 0s 4ms/step - loss: 0.5883 - accuracy: 0.5807 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 24/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5158 - accuracy: 0.5938 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 25/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5409 - accuracy: 0.6157 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 26/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5141 - accuracy: 0.6063 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 27/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4711 - accuracy: 0.6110 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 28/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5177 - accuracy: 0.5800 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 29/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4413 - accuracy: 0.6290 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 30/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4567 - accuracy: 0.6199 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 31/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4971 - accuracy: 0.5711 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 32/100
42/42 [==============================] - 0s 6ms/step - loss: 0.4004 - accuracy: 0.6516 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 33/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5323 - accuracy: 0.6022 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 34/100
42/42 [==============================] - 0s 6ms/step - loss: 0.4825 - accuracy: 0.6244 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 35/100
42/42 [==============================] - 1s 13ms/step - loss: 0.4786 - accuracy: 0.5561 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 36/100
42/42 [==============================] - 0s 7ms/step - loss: 0.4848 - accuracy: 0.6041 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 37/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5379 - accuracy: 0.6158 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 38/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5322 - accuracy: 0.5951 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 39/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5593 - accuracy: 0.5829 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 40/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5847 - accuracy: 0.5470 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 41/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5604 - accuracy: 0.5886 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 42/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5913 - accuracy: 0.5788 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 43/100
42/42 [==============================] - 0s 6ms/step - loss: 0.4943 - accuracy: 0.6185 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 44/100
42/42 [==============================] - 0s 5ms/step - loss: 0.3988 - accuracy: 0.6362 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 45/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5136 - accuracy: 0.5918 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 46/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5704 - accuracy: 0.5944 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 47/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5763 - accuracy: 0.6094 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 48/100
42/42 [==============================] - 1s 17ms/step - loss: 0.5097 - accuracy: 0.5769 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 49/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5034 - accuracy: 0.5806 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 50/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5289 - accuracy: 0.5853 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 51/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4845 - accuracy: 0.5917 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 52/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4805 - accuracy: 0.6164 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 53/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5653 - accuracy: 0.6009 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 54/100
42/42 [==============================] - 0s 8ms/step - loss: 0.5218 - accuracy: 0.5788 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 55/100
42/42 [==============================] - 0s 6ms/step - loss: 0.4258 - accuracy: 0.6114 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 56/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5298 - accuracy: 0.6094 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 57/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4984 - accuracy: 0.5897 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 58/100
42/42 [==============================] - 0s 4ms/step - loss: 0.5610 - accuracy: 0.6227 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 59/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4819 - accuracy: 0.6228 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 60/100
42/42 [==============================] - 0s 10ms/step - loss: 0.5598 - accuracy: 0.5810 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 61/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4780 - accuracy: 0.6249 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 62/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5722 - accuracy: 0.6010 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 63/100
42/42 [==============================] - 0s 9ms/step - loss: 0.4778 - accuracy: 0.5815 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 64/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5505 - accuracy: 0.5736 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 65/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5449 - accuracy: 0.6031 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 66/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5663 - accuracy: 0.5759 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 67/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5102 - accuracy: 0.6178 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 68/100
42/42 [==============================] - 1s 13ms/step - loss: 0.5173 - accuracy: 0.5865 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 69/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5516 - accuracy: 0.6023 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 70/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4825 - accuracy: 0.6482 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 71/100
42/42 [==============================] - 0s 10ms/step - loss: 0.5273 - accuracy: 0.6095 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 72/100
42/42 [==============================] - 1s 15ms/step - loss: 0.5345 - accuracy: 0.6161 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 73/100
42/42 [==============================] - 0s 9ms/step - loss: 0.5085 - accuracy: 0.6090 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 74/100
42/42 [==============================] - 0s 8ms/step - loss: 0.4524 - accuracy: 0.6330 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 75/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4541 - accuracy: 0.6217 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 76/100
42/42 [==============================] - 0s 6ms/step - loss: 0.4838 - accuracy: 0.6187 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 77/100
42/42 [==============================] - 0s 6ms/step - loss: 0.4772 - accuracy: 0.6010 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 78/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5533 - accuracy: 0.6375 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 79/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4865 - accuracy: 0.6276 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 80/100
42/42 [==============================] - 0s 5ms/step - loss: 0.4844 - accuracy: 0.6060 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 81/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5111 - accuracy: 0.6106 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 82/100
42/42 [==============================] - 0s 5ms/step - loss: 0.5322 - accuracy: 0.5927 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 83/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5010 - accuracy: 0.6210 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 84/100
42/42 [==============================] - 0s 10ms/step - loss: 0.4069 - accuracy: 0.6066 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 85/100
42/42 [==============================] - 0s 6ms/step - loss: 0.4869 - accuracy: 0.5643 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 86/100
42/42 [==============================] - 0s 6ms/step - loss: 0.4902 - accuracy: 0.5856 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 87/100
42/42 [==============================] - 1s 15ms/step - loss: 0.5199 - accuracy: 0.6158 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 88/100
42/42 [==============================] - 0s 8ms/step - loss: 0.4625 - accuracy: 0.6091 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 89/100
42/42 [==============================] - 0s 9ms/step - loss: 0.4256 - accuracy: 0.6248 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 90/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5048 - accuracy: 0.5788 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 91/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5264 - accuracy: 0.5721 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 92/100
42/42 [==============================] - 0s 9ms/step - loss: 0.5467 - accuracy: 0.5950 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 93/100
42/42 [==============================] - 0s 7ms/step - loss: 0.5022 - accuracy: 0.6299 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 94/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5267 - accuracy: 0.6035 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 95/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5707 - accuracy: 0.5857 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 96/100
42/42 [==============================] - 0s 8ms/step - loss: 0.4532 - accuracy: 0.6142 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 97/100
42/42 [==============================] - 0s 11ms/step - loss: 0.5340 - accuracy: 0.5847 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 98/100
42/42 [==============================] - 0s 7ms/step - loss: 0.5338 - accuracy: 0.5988 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 99/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5599 - accuracy: 0.5808 - val_loss: 0.9905 - val_accuracy: 0.5000
Epoch 100/100
42/42 [==============================] - 0s 6ms/step - loss: 0.5624 - accuracy: 0.5960 - val_loss: 0.9905 - val_accuracy: 0.5000

model.summary()

Model: "sequential"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
dense (Dense)                (None, 50)                2150      
_________________________________________________________________
dense_1 (Dense)              (None, 100)               5100      
_________________________________________________________________
dense_2 (Dense)              (None, 150)               15150     
_________________________________________________________________
dense_3 (Dense)              (None, 40)                6040      
_________________________________________________________________
dense_4 (Dense)              (None, 1)                 41        
=================================================================
Total params: 28,481
Trainable params: 28,481
Non-trainable params: 0
_________________________________________________________________

# evaluate the keras model
_, train_accuracy = model.evaluate(X_train, y_train, batch_size=8, verbose=0)
_, test_accuracy = model.evaluate(X_test, y_test, batch_size=8, verbose=0)

print(color.GREEN + 'Train accuracy: ' +color.END + str(train_accuracy*100))
print(color.GREEN + 'Test accuracy: ' + color.END + str(test_accuracy*100))

Train accuracy: 60.17963886260986
Test accuracy: 50.0

accuracy, precision, recall, f1 = get_classification_metrics(model, X_test, y_test, '')
print(color.BLUE + 'Accuracy: ' + color.END + str(accuracy))
print(color.BLUE + 'Precision: ' + color.END + str(precision))
print(color.BLUE + 'Recall: ' + color.END +  str(recall))
print(color.BLUE + 'F1 score: ' + color.END + str(f1))

Accuracy: 0.4880952380952381
Precision: 0.4880952380952381
Recall: 0.4880952380952381
F1 score: 0.4880952380952381

epochs = range(len(training_history.history['loss'])) # Get number of epochs

# plot loss learning curves
plt.plot  (epochs, training_history.history['loss'], label = 'train')
plt.plot  (epochs, training_history.history['val_loss'], label = 'val')
plt.legend(loc = 'upper right')
plt.title ('Training and validation loss')

Text(0.5,1,'Training and validation loss')

# plot accuracy learning curves
plt.plot(epochs, training_history.history['accuracy'], label = 'train')
plt.plot(epochs, training_history.history['val_accuracy'], label = 'test')
plt.legend(loc = 'upper right')
plt.title ('Training and validation accuracy')

Text(0.5,1,'Training and validation accuracy')

Multiclass Classification - Data with One-hot Encoded Target variable¶

random_state = 42
#param = 1e-9
param = 1e-4

# define the model
model = Sequential()

model.add(Dense(10, input_dim=X_train.shape[1], activation='relu', kernel_initializer='he_uniform', kernel_regularizer=l2(param),
                kernel_constraint=unit_norm()))
model.add(Dropout(0.2))
model.add(BatchNormalization())
model.add(Dense(10, activation='relu', kernel_initializer='he_uniform', kernel_regularizer=l2(param), 
                kernel_constraint=unit_norm()))
model.add(Dropout(0.5))
model.add(BatchNormalization())
model.add(Dense(5, activation='softmax', kernel_regularizer=l2(param), 
                kernel_constraint=unit_norm())) # Multilabel

# compile the keras model
#opt = optimizers.Adamax(lr=0.01)
opt = SGD(lr=0.001, momentum=0.9)
model.compile(loss='categorical_crossentropy', optimizer=opt, metrics=['categorical_accuracy'])

# Use earlystopping
callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=2, min_delta=1E-3)
rlrp = ReduceLROnPlateau(monitor='val_loss', factor=0.0001, patience=2, min_delta=1E-4)

target_type = 'multi_label'
metrics = Metrics(validation_data=(X_train, y_train_dummy, target_type))

# fit the keras model on the dataset
training_history = model.fit(X_train, y_train_dummy, epochs=30, batch_size=8, verbose=1, validation_data=(X_test, y_test_dummy), callbacks=[rlrp, metrics])

Epoch 1/30
42/42 [==============================] - 2s 15ms/step - loss: 1.8607 - categorical_accuracy: 0.2621 - val_loss: 1.4091 - val_categorical_accuracy: 0.5357
Epoch 2/30
42/42 [==============================] - 0s 6ms/step - loss: 1.5224 - categorical_accuracy: 0.4636 - val_loss: 1.1892 - val_categorical_accuracy: 0.6667
Epoch 3/30
42/42 [==============================] - 0s 4ms/step - loss: 1.3699 - categorical_accuracy: 0.5326 - val_loss: 1.0751 - val_categorical_accuracy: 0.7024
Epoch 4/30
42/42 [==============================] - 0s 4ms/step - loss: 1.3232 - categorical_accuracy: 0.5730 - val_loss: 1.0067 - val_categorical_accuracy: 0.7143
Epoch 5/30
42/42 [==============================] - 0s 8ms/step - loss: 1.1348 - categorical_accuracy: 0.6409 - val_loss: 0.9727 - val_categorical_accuracy: 0.7143
Epoch 6/30
42/42 [==============================] - 0s 4ms/step - loss: 1.1459 - categorical_accuracy: 0.6720 - val_loss: 0.9536 - val_categorical_accuracy: 0.7381
Epoch 7/30
42/42 [==============================] - 0s 8ms/step - loss: 1.0472 - categorical_accuracy: 0.7377 - val_loss: 0.9514 - val_categorical_accuracy: 0.7381
Epoch 8/30
42/42 [==============================] - 0s 4ms/step - loss: 1.0298 - categorical_accuracy: 0.7079 - val_loss: 0.9537 - val_categorical_accuracy: 0.7381
Epoch 9/30
42/42 [==============================] - 0s 8ms/step - loss: 1.0466 - categorical_accuracy: 0.7226 - val_loss: 0.9551 - val_categorical_accuracy: 0.7381
Epoch 10/30
42/42 [==============================] - 0s 5ms/step - loss: 0.8731 - categorical_accuracy: 0.7580 - val_loss: 0.9509 - val_categorical_accuracy: 0.7381
Epoch 11/30
42/42 [==============================] - 0s 4ms/step - loss: 1.0977 - categorical_accuracy: 0.7023 - val_loss: 0.9547 - val_categorical_accuracy: 0.7381
Epoch 12/30
42/42 [==============================] - 0s 4ms/step - loss: 0.9977 - categorical_accuracy: 0.7280 - val_loss: 0.9524 - val_categorical_accuracy: 0.7381
Epoch 13/30
42/42 [==============================] - 0s 4ms/step - loss: 0.9503 - categorical_accuracy: 0.7266 - val_loss: 0.9515 - val_categorical_accuracy: 0.7381
Epoch 14/30
42/42 [==============================] - 0s 4ms/step - loss: 0.9510 - categorical_accuracy: 0.7372 - val_loss: 0.9510 - val_categorical_accuracy: 0.7381
Epoch 15/30
42/42 [==============================] - 0s 4ms/step - loss: 0.9791 - categorical_accuracy: 0.7327 - val_loss: 0.9519 - val_categorical_accuracy: 0.7381
Epoch 16/30
42/42 [==============================] - 0s 5ms/step - loss: 1.0256 - categorical_accuracy: 0.7448 - val_loss: 0.9519 - val_categorical_accuracy: 0.7381
Epoch 17/30
42/42 [==============================] - 0s 5ms/step - loss: 0.9367 - categorical_accuracy: 0.7374 - val_loss: 0.9530 - val_categorical_accuracy: 0.7381
Epoch 18/30
42/42 [==============================] - 0s 5ms/step - loss: 1.0497 - categorical_accuracy: 0.7212 - val_loss: 0.9529 - val_categorical_accuracy: 0.7381
Epoch 19/30
42/42 [==============================] - 0s 5ms/step - loss: 0.9390 - categorical_accuracy: 0.7655 - val_loss: 0.9494 - val_categorical_accuracy: 0.7381
Epoch 20/30
42/42 [==============================] - 0s 5ms/step - loss: 1.0597 - categorical_accuracy: 0.7032 - val_loss: 0.9505 - val_categorical_accuracy: 0.7381
Epoch 21/30
42/42 [==============================] - 0s 6ms/step - loss: 1.0257 - categorical_accuracy: 0.7257 - val_loss: 0.9549 - val_categorical_accuracy: 0.7381
Epoch 22/30
42/42 [==============================] - 0s 8ms/step - loss: 1.1394 - categorical_accuracy: 0.6682 - val_loss: 0.9540 - val_categorical_accuracy: 0.7381
Epoch 23/30
42/42 [==============================] - 0s 4ms/step - loss: 1.0961 - categorical_accuracy: 0.6855 - val_loss: 0.9503 - val_categorical_accuracy: 0.7381
Epoch 24/30
42/42 [==============================] - 0s 4ms/step - loss: 0.9629 - categorical_accuracy: 0.7487 - val_loss: 0.9492 - val_categorical_accuracy: 0.7381
Epoch 25/30
42/42 [==============================] - 0s 4ms/step - loss: 0.9582 - categorical_accuracy: 0.7398 - val_loss: 0.9501 - val_categorical_accuracy: 0.7381
Epoch 26/30
42/42 [==============================] - 0s 4ms/step - loss: 0.9475 - categorical_accuracy: 0.7545 - val_loss: 0.9536 - val_categorical_accuracy: 0.7381
Epoch 27/30
42/42 [==============================] - 0s 5ms/step - loss: 1.0967 - categorical_accuracy: 0.7014 - val_loss: 0.9533 - val_categorical_accuracy: 0.7381
Epoch 28/30
42/42 [==============================] - 0s 5ms/step - loss: 0.9772 - categorical_accuracy: 0.7351 - val_loss: 0.9512 - val_categorical_accuracy: 0.7381
Epoch 29/30
42/42 [==============================] - 0s 5ms/step - loss: 1.0282 - categorical_accuracy: 0.7373 - val_loss: 0.9520 - val_categorical_accuracy: 0.7381
Epoch 30/30
42/42 [==============================] - 0s 6ms/step - loss: 1.0188 - categorical_accuracy: 0.7410 - val_loss: 0.9544 - val_categorical_accuracy: 0.7381

model.summary()

Model: "sequential_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
dense_5 (Dense)              (None, 10)                430       
_________________________________________________________________
dropout (Dropout)            (None, 10)                0         
_________________________________________________________________
batch_normalization (BatchNo (None, 10)                40        
_________________________________________________________________
dense_6 (Dense)              (None, 10)                110       
_________________________________________________________________
dropout_1 (Dropout)          (None, 10)                0         
_________________________________________________________________
batch_normalization_1 (Batch (None, 10)                40        
_________________________________________________________________
dense_7 (Dense)              (None, 5)                 55        
=================================================================
Total params: 675
Trainable params: 635
Non-trainable params: 40
_________________________________________________________________

# evaluate the keras model
_, train_accuracy = model.evaluate(X_train, y_train_dummy, batch_size=8, verbose=0)
_, test_accuracy = model.evaluate(X_test, y_test_dummy, batch_size=8, verbose=0)

print(color.GREEN + 'Train accuracy: ' +color.END + str(train_accuracy*100))
print(color.GREEN + 'Test accuracy: ' + color.END + str(test_accuracy*100))

Train accuracy: 73.35329055786133
Test accuracy: 73.8095223903656

accuracy, precision, recall, f1 = get_classification_metrics(model, X_test, y_test_dummy, target_type)
print(color.BLUE + 'Accuracy: ' + color.END + str(accuracy))
print(color.BLUE + 'Precision: ' + color.END + str(precision))
print(color.BLUE + 'Recall: ' + color.END +  str(recall))
print(color.BLUE + 'F1 score: ' + color.END + str(f1))

Accuracy: 0.7023809523809523
Precision: 0.7283950617283951
Recall: 0.7023809523809523
F1 score: 0.7151515151515152

epochs = range(len(training_history.history['loss'])) # Get number of epochs

# plot loss learning curves
plt.plot(epochs, training_history.history['loss'], label = 'train')
plt.plot(epochs, training_history.history['val_loss'], label = 'test')
plt.legend(loc = 'upper right')
plt.title ('Training and validation loss')

Text(0.5,1,'Training and validation loss')

# plot accuracy learning curves
plt.plot(epochs, training_history.history['categorical_accuracy'], label = 'train')
plt.plot(epochs, training_history.history['val_categorical_accuracy'], label = 'test')
plt.legend(loc = 'upper right')
plt.title ('Training and validation accuracy')

Text(0.5,1,'Training and validation accuracy')

# serialize model to JSON
model_json = model.to_json()
with open("model.json", "w") as json_file:
    json_file.write(model_json)
    
# serialize weights to HDF5
model.save_weights("model.h5")
print(color.GREEN + "Saved model weights to disk" + color.END)

# Save the model in h5 format 
model.save("finalized_keras_model.h5")
print(color.GREEN + "Saved model to disk" + color.END)

Saved model weights to disk
Saved model to disk

Multiclass Classification - Data with SMOTE and One-hot Encoded Target variable¶

# fix random seed for reproducibility
random_state = 42
#param = 1e-9
param = 1e-4

# define the model
model = Sequential()

model.add(Dense(10, input_dim=X_train_smote.shape[1], activation='relu', kernel_initializer='he_uniform', kernel_regularizer=l2(param),
                kernel_constraint=unit_norm()))
model.add(Dropout(0.2))
model.add(BatchNormalization())
model.add(Dense(10, activation='relu', kernel_initializer='he_uniform', kernel_regularizer=l2(param), 
                kernel_constraint=unit_norm()))
model.add(Dropout(0.5))
model.add(BatchNormalization())
model.add(Dense(5, activation='softmax', kernel_regularizer=l2(param), 
                kernel_constraint=unit_norm())) # Multilabel

# compile the keras model
#opt = optimizers.Adamax(lr=0.01)
opt = SGD(lr=0.001, momentum=0.9)
model.compile(loss='categorical_crossentropy', optimizer=opt, metrics=['categorical_accuracy'])

# Use earlystopping
callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=7, min_delta=1E-3)
rlrp = ReduceLROnPlateau(monitor='val_loss', factor=0.0001, patience=5, min_delta=1E-4)

target_type = 'multi_label'
metrics = Metrics(validation_data=(X_train_smote, y_train_smote_dummy, target_type))

# fit the keras model on the dataset
training_history = model.fit(X_train_smote, y_train_smote_dummy, epochs=100, batch_size=8, verbose=1, validation_data=(X_test, y_test_dummy), callbacks=[rlrp, metrics])

Epoch 1/100
155/155 [==============================] - 2s 5ms/step - loss: 1.9103 - categorical_accuracy: 0.2070 - val_loss: 1.6795 - val_categorical_accuracy: 0.1190
Epoch 2/100
155/155 [==============================] - 0s 3ms/step - loss: 1.6965 - categorical_accuracy: 0.2919 - val_loss: 1.7077 - val_categorical_accuracy: 0.1190
Epoch 3/100
155/155 [==============================] - 1s 4ms/step - loss: 1.6039 - categorical_accuracy: 0.3190 - val_loss: 1.6507 - val_categorical_accuracy: 0.1905
Epoch 4/100
155/155 [==============================] - 1s 7ms/step - loss: 1.5982 - categorical_accuracy: 0.2999 - val_loss: 1.6349 - val_categorical_accuracy: 0.2381
Epoch 5/100
155/155 [==============================] - 1s 3ms/step - loss: 1.5125 - categorical_accuracy: 0.3626 - val_loss: 1.6364 - val_categorical_accuracy: 0.2381
Epoch 6/100
155/155 [==============================] - 1s 3ms/step - loss: 1.5431 - categorical_accuracy: 0.3223 - val_loss: 1.6248 - val_categorical_accuracy: 0.2500
Epoch 7/100
155/155 [==============================] - 1s 3ms/step - loss: 1.5147 - categorical_accuracy: 0.3223 - val_loss: 1.6196 - val_categorical_accuracy: 0.2262
Epoch 8/100
155/155 [==============================] - 0s 2ms/step - loss: 1.5247 - categorical_accuracy: 0.3180 - val_loss: 1.5939 - val_categorical_accuracy: 0.1905
Epoch 9/100
155/155 [==============================] - 0s 3ms/step - loss: 1.5112 - categorical_accuracy: 0.3146 - val_loss: 1.5869 - val_categorical_accuracy: 0.1786
Epoch 10/100
155/155 [==============================] - 0s 3ms/step - loss: 1.5156 - categorical_accuracy: 0.3116 - val_loss: 1.5876 - val_categorical_accuracy: 0.1786
Epoch 11/100
155/155 [==============================] - 1s 4ms/step - loss: 1.4745 - categorical_accuracy: 0.3458 - val_loss: 1.5924 - val_categorical_accuracy: 0.1905
Epoch 12/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4266 - categorical_accuracy: 0.3887 - val_loss: 1.6197 - val_categorical_accuracy: 0.1786
Epoch 13/100
155/155 [==============================] - 0s 2ms/step - loss: 1.4678 - categorical_accuracy: 0.3413 - val_loss: 1.5897 - val_categorical_accuracy: 0.1548
Epoch 14/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4533 - categorical_accuracy: 0.3719 - val_loss: 1.5832 - val_categorical_accuracy: 0.1905
Epoch 15/100
155/155 [==============================] - 0s 2ms/step - loss: 1.4287 - categorical_accuracy: 0.3622 - val_loss: 1.5759 - val_categorical_accuracy: 0.1548
Epoch 16/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4229 - categorical_accuracy: 0.3752 - val_loss: 1.5416 - val_categorical_accuracy: 0.2143
Epoch 17/100
155/155 [==============================] - 1s 5ms/step - loss: 1.4236 - categorical_accuracy: 0.3675 - val_loss: 1.5348 - val_categorical_accuracy: 0.2024
Epoch 18/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4149 - categorical_accuracy: 0.3825 - val_loss: 1.5802 - val_categorical_accuracy: 0.1786
Epoch 19/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4302 - categorical_accuracy: 0.3663 - val_loss: 1.5833 - val_categorical_accuracy: 0.1667
Epoch 20/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4103 - categorical_accuracy: 0.3926 - val_loss: 1.5835 - val_categorical_accuracy: 0.1548
Epoch 21/100
155/155 [==============================] - 1s 5ms/step - loss: 1.3886 - categorical_accuracy: 0.3834 - val_loss: 1.5584 - val_categorical_accuracy: 0.1786
Epoch 22/100
155/155 [==============================] - 1s 3ms/step - loss: 1.3492 - categorical_accuracy: 0.4036 - val_loss: 1.5589 - val_categorical_accuracy: 0.1905
Epoch 23/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4033 - categorical_accuracy: 0.4100 - val_loss: 1.5948 - val_categorical_accuracy: 0.1786
Epoch 24/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4201 - categorical_accuracy: 0.3700 - val_loss: 1.6148 - val_categorical_accuracy: 0.1667
Epoch 25/100
155/155 [==============================] - 1s 3ms/step - loss: 1.3847 - categorical_accuracy: 0.4096 - val_loss: 1.6178 - val_categorical_accuracy: 0.1548
Epoch 26/100
155/155 [==============================] - 1s 5ms/step - loss: 1.3825 - categorical_accuracy: 0.3961 - val_loss: 1.5985 - val_categorical_accuracy: 0.1905
Epoch 27/100
155/155 [==============================] - 1s 9ms/step - loss: 1.3971 - categorical_accuracy: 0.3794 - val_loss: 1.5953 - val_categorical_accuracy: 0.1905
Epoch 28/100
155/155 [==============================] - 1s 5ms/step - loss: 1.4081 - categorical_accuracy: 0.3806 - val_loss: 1.6020 - val_categorical_accuracy: 0.1905
Epoch 29/100
155/155 [==============================] - 1s 5ms/step - loss: 1.3719 - categorical_accuracy: 0.3623 - val_loss: 1.6043 - val_categorical_accuracy: 0.1786
Epoch 30/100
155/155 [==============================] - 1s 8ms/step - loss: 1.4086 - categorical_accuracy: 0.3999 - val_loss: 1.5982 - val_categorical_accuracy: 0.1667
Epoch 31/100
155/155 [==============================] - 1s 5ms/step - loss: 1.3999 - categorical_accuracy: 0.3674 - val_loss: 1.6063 - val_categorical_accuracy: 0.1786
Epoch 32/100
155/155 [==============================] - 1s 5ms/step - loss: 1.4080 - categorical_accuracy: 0.3906 - val_loss: 1.6111 - val_categorical_accuracy: 0.1548
Epoch 33/100
155/155 [==============================] - 1s 5ms/step - loss: 1.3692 - categorical_accuracy: 0.4068 - val_loss: 1.5954 - val_categorical_accuracy: 0.1786
Epoch 34/100
155/155 [==============================] - 1s 5ms/step - loss: 1.3796 - categorical_accuracy: 0.4089 - val_loss: 1.6043 - val_categorical_accuracy: 0.1667
Epoch 35/100
155/155 [==============================] - 1s 8ms/step - loss: 1.3932 - categorical_accuracy: 0.3751 - val_loss: 1.6067 - val_categorical_accuracy: 0.1667
Epoch 36/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3957 - categorical_accuracy: 0.3980 - val_loss: 1.6159 - val_categorical_accuracy: 0.1667
Epoch 37/100
155/155 [==============================] - 1s 6ms/step - loss: 1.3605 - categorical_accuracy: 0.4231 - val_loss: 1.6030 - val_categorical_accuracy: 0.1667
Epoch 38/100
155/155 [==============================] - 1s 4ms/step - loss: 1.4100 - categorical_accuracy: 0.3741 - val_loss: 1.6060 - val_categorical_accuracy: 0.1667
Epoch 39/100
155/155 [==============================] - 1s 3ms/step - loss: 1.3887 - categorical_accuracy: 0.3850 - val_loss: 1.5996 - val_categorical_accuracy: 0.1667
Epoch 40/100
155/155 [==============================] - 1s 5ms/step - loss: 1.4078 - categorical_accuracy: 0.3828 - val_loss: 1.6105 - val_categorical_accuracy: 0.1667
Epoch 41/100
155/155 [==============================] - 1s 6ms/step - loss: 1.4013 - categorical_accuracy: 0.3842 - val_loss: 1.6086 - val_categorical_accuracy: 0.1548
Epoch 42/100
155/155 [==============================] - 1s 7ms/step - loss: 1.3741 - categorical_accuracy: 0.4075 - val_loss: 1.6034 - val_categorical_accuracy: 0.1786
Epoch 43/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3843 - categorical_accuracy: 0.3768 - val_loss: 1.6061 - val_categorical_accuracy: 0.1786
Epoch 44/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3853 - categorical_accuracy: 0.4108 - val_loss: 1.5906 - val_categorical_accuracy: 0.1786
Epoch 45/100
155/155 [==============================] - 1s 3ms/step - loss: 1.3788 - categorical_accuracy: 0.4043 - val_loss: 1.5954 - val_categorical_accuracy: 0.1786
Epoch 46/100
155/155 [==============================] - 1s 4ms/step - loss: 1.4150 - categorical_accuracy: 0.3729 - val_loss: 1.6072 - val_categorical_accuracy: 0.1786
Epoch 47/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3808 - categorical_accuracy: 0.4471 - val_loss: 1.6127 - val_categorical_accuracy: 0.1667
Epoch 48/100
155/155 [==============================] - 1s 4ms/step - loss: 1.4180 - categorical_accuracy: 0.3655 - val_loss: 1.6067 - val_categorical_accuracy: 0.1786
Epoch 49/100
155/155 [==============================] - 1s 3ms/step - loss: 1.3519 - categorical_accuracy: 0.4171 - val_loss: 1.6136 - val_categorical_accuracy: 0.1786
Epoch 50/100
155/155 [==============================] - 1s 5ms/step - loss: 1.4000 - categorical_accuracy: 0.3761 - val_loss: 1.6205 - val_categorical_accuracy: 0.1667
Epoch 51/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3772 - categorical_accuracy: 0.4298 - val_loss: 1.5977 - val_categorical_accuracy: 0.1905
Epoch 52/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3567 - categorical_accuracy: 0.4113 - val_loss: 1.6061 - val_categorical_accuracy: 0.1786
Epoch 53/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4049 - categorical_accuracy: 0.3833 - val_loss: 1.6057 - val_categorical_accuracy: 0.1905
Epoch 54/100
155/155 [==============================] - 1s 4ms/step - loss: 1.4057 - categorical_accuracy: 0.3992 - val_loss: 1.6212 - val_categorical_accuracy: 0.1786
Epoch 55/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3747 - categorical_accuracy: 0.4031 - val_loss: 1.6213 - val_categorical_accuracy: 0.1667
Epoch 56/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3827 - categorical_accuracy: 0.3935 - val_loss: 1.6204 - val_categorical_accuracy: 0.1667
Epoch 57/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3930 - categorical_accuracy: 0.3785 - val_loss: 1.6055 - val_categorical_accuracy: 0.1786
Epoch 58/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3862 - categorical_accuracy: 0.4078 - val_loss: 1.6194 - val_categorical_accuracy: 0.1548
Epoch 59/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4097 - categorical_accuracy: 0.3873 - val_loss: 1.6001 - val_categorical_accuracy: 0.1667
Epoch 60/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4076 - categorical_accuracy: 0.4108 - val_loss: 1.6068 - val_categorical_accuracy: 0.1667
Epoch 61/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3523 - categorical_accuracy: 0.4337 - val_loss: 1.5913 - val_categorical_accuracy: 0.1786
Epoch 62/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3797 - categorical_accuracy: 0.3907 - val_loss: 1.5997 - val_categorical_accuracy: 0.1905
Epoch 63/100
155/155 [==============================] - 1s 3ms/step - loss: 1.3812 - categorical_accuracy: 0.4092 - val_loss: 1.6033 - val_categorical_accuracy: 0.1786
Epoch 64/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4059 - categorical_accuracy: 0.3885 - val_loss: 1.6020 - val_categorical_accuracy: 0.1786
Epoch 65/100
155/155 [==============================] - 1s 3ms/step - loss: 1.4089 - categorical_accuracy: 0.3696 - val_loss: 1.6000 - val_categorical_accuracy: 0.1786
Epoch 66/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3985 - categorical_accuracy: 0.3847 - val_loss: 1.6132 - val_categorical_accuracy: 0.1667
Epoch 67/100
155/155 [==============================] - 1s 5ms/step - loss: 1.4167 - categorical_accuracy: 0.3868 - val_loss: 1.6021 - val_categorical_accuracy: 0.1667
Epoch 68/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4008 - categorical_accuracy: 0.3879 - val_loss: 1.6051 - val_categorical_accuracy: 0.1786
Epoch 69/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3962 - categorical_accuracy: 0.4034 - val_loss: 1.6048 - val_categorical_accuracy: 0.1667
Epoch 70/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3920 - categorical_accuracy: 0.3762 - val_loss: 1.6092 - val_categorical_accuracy: 0.1667
Epoch 71/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4133 - categorical_accuracy: 0.3732 - val_loss: 1.5964 - val_categorical_accuracy: 0.1786
Epoch 72/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3853 - categorical_accuracy: 0.3939 - val_loss: 1.5942 - val_categorical_accuracy: 0.1905
Epoch 73/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4048 - categorical_accuracy: 0.3626 - val_loss: 1.5979 - val_categorical_accuracy: 0.1905
Epoch 74/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4221 - categorical_accuracy: 0.3625 - val_loss: 1.5947 - val_categorical_accuracy: 0.1905
Epoch 75/100
155/155 [==============================] - 1s 3ms/step - loss: 1.3857 - categorical_accuracy: 0.4023 - val_loss: 1.6035 - val_categorical_accuracy: 0.1786
Epoch 76/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3993 - categorical_accuracy: 0.3971 - val_loss: 1.5899 - val_categorical_accuracy: 0.1905
Epoch 77/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3907 - categorical_accuracy: 0.3807 - val_loss: 1.6011 - val_categorical_accuracy: 0.1905
Epoch 78/100
155/155 [==============================] - 1s 6ms/step - loss: 1.3805 - categorical_accuracy: 0.4139 - val_loss: 1.6052 - val_categorical_accuracy: 0.1667
Epoch 79/100
155/155 [==============================] - 1s 6ms/step - loss: 1.3626 - categorical_accuracy: 0.4071 - val_loss: 1.6136 - val_categorical_accuracy: 0.1786
Epoch 80/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4077 - categorical_accuracy: 0.3974 - val_loss: 1.6062 - val_categorical_accuracy: 0.1905
Epoch 81/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3717 - categorical_accuracy: 0.3935 - val_loss: 1.6056 - val_categorical_accuracy: 0.1786
Epoch 82/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4008 - categorical_accuracy: 0.3892 - val_loss: 1.6063 - val_categorical_accuracy: 0.1667
Epoch 83/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4314 - categorical_accuracy: 0.3945 - val_loss: 1.6039 - val_categorical_accuracy: 0.1786
Epoch 84/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3776 - categorical_accuracy: 0.3929 - val_loss: 1.6054 - val_categorical_accuracy: 0.1905
Epoch 85/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3952 - categorical_accuracy: 0.3879 - val_loss: 1.5931 - val_categorical_accuracy: 0.1905
Epoch 86/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4223 - categorical_accuracy: 0.3973 - val_loss: 1.5978 - val_categorical_accuracy: 0.1786
Epoch 87/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4057 - categorical_accuracy: 0.3920 - val_loss: 1.6110 - val_categorical_accuracy: 0.1667
Epoch 88/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3788 - categorical_accuracy: 0.4227 - val_loss: 1.6060 - val_categorical_accuracy: 0.1905
Epoch 89/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3657 - categorical_accuracy: 0.3866 - val_loss: 1.6108 - val_categorical_accuracy: 0.1786
Epoch 90/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3681 - categorical_accuracy: 0.4137 - val_loss: 1.5950 - val_categorical_accuracy: 0.1667
Epoch 91/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3541 - categorical_accuracy: 0.4398 - val_loss: 1.6096 - val_categorical_accuracy: 0.1667
Epoch 92/100
155/155 [==============================] - 1s 4ms/step - loss: 1.4118 - categorical_accuracy: 0.3921 - val_loss: 1.6149 - val_categorical_accuracy: 0.1667
Epoch 93/100
155/155 [==============================] - 0s 3ms/step - loss: 1.4230 - categorical_accuracy: 0.3842 - val_loss: 1.5942 - val_categorical_accuracy: 0.1905
Epoch 94/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3915 - categorical_accuracy: 0.4015 - val_loss: 1.6124 - val_categorical_accuracy: 0.1667
Epoch 95/100
155/155 [==============================] - 0s 3ms/step - loss: 1.3853 - categorical_accuracy: 0.4116 - val_loss: 1.6087 - val_categorical_accuracy: 0.1786
Epoch 96/100
155/155 [==============================] - 1s 3ms/step - loss: 1.3686 - categorical_accuracy: 0.4108 - val_loss: 1.6070 - val_categorical_accuracy: 0.1786
Epoch 97/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3840 - categorical_accuracy: 0.4107 - val_loss: 1.6043 - val_categorical_accuracy: 0.1905
Epoch 98/100
155/155 [==============================] - 1s 3ms/step - loss: 1.3787 - categorical_accuracy: 0.3762 - val_loss: 1.6003 - val_categorical_accuracy: 0.1905
Epoch 99/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3992 - categorical_accuracy: 0.3727 - val_loss: 1.6069 - val_categorical_accuracy: 0.1905
Epoch 100/100
155/155 [==============================] - 1s 4ms/step - loss: 1.3965 - categorical_accuracy: 0.4079 - val_loss: 1.5944 - val_categorical_accuracy: 0.1905

model.summary()

Model: "sequential_2"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
dense_8 (Dense)              (None, 10)                430       
_________________________________________________________________
dropout_2 (Dropout)          (None, 10)                0         
_________________________________________________________________
batch_normalization_2 (Batch (None, 10)                40        
_________________________________________________________________
dense_9 (Dense)              (None, 10)                110       
_________________________________________________________________
dropout_3 (Dropout)          (None, 10)                0         
_________________________________________________________________
batch_normalization_3 (Batch (None, 10)                40        
_________________________________________________________________
dense_10 (Dense)             (None, 5)                 55        
=================================================================
Total params: 675
Trainable params: 635
Non-trainable params: 40
_________________________________________________________________

# evaluate the keras model
_, train_accuracy = model.evaluate(X_train_smote, y_train_smote_dummy, batch_size=8, verbose=0)
_, test_accuracy = model.evaluate(X_test, y_test_dummy, batch_size=8, verbose=0)

print(color.GREEN + 'Train accuracy: ' + color.END + str(train_accuracy*100))
print(color.GREEN + 'Test accuracy: ' + color.END + str(test_accuracy*100))

Train accuracy: 52.06477642059326
Test accuracy: 19.0476194024086

accuracy, precision, recall, f1 = get_classification_metrics(model, X_test, y_test_dummy, target_type)
print(color.BLUE + 'Accuracy: ' + color.END + str(accuracy))
print(color.BLUE + 'Precision: ' + color.END + str(precision))
print(color.BLUE + 'Recall: ' + color.END + str(recall))
print(color.BLUE + 'F1 score: ' + color.END + str(f1))

Accuracy: 0.011904761904761904
Precision: 0.14285714285714285
Recall: 0.011904761904761904
F1 score: 0.021978021978021973

epochs = range(len(training_history.history['loss'])) # Get number of epochs

# plot loss learning curves
plt.plot(epochs, training_history.history['loss'], label = 'train')
plt.plot(epochs, training_history.history['val_loss'], label = 'test')
plt.legend(loc = 'upper right')
plt.title ('Training and validation loss')

Text(0.5,1,'Training and validation loss')

# plot accuracy learning curves
plt.plot(epochs, training_history.history['categorical_accuracy'], label = 'train')
plt.plot(epochs, training_history.history['val_categorical_accuracy'], label = 'test')
plt.legend(loc = 'upper right')
plt.title ('Training and validation accuracy')

Text(0.5,1,'Training and validation accuracy')

# serialize model to JSON
model_json = model.to_json()
with open("model.json", "w") as json_file:
    json_file.write(model_json)
    
# serialize weights to HDF5
model.save_weights("model.h5")
print(color.GREEN + "Saved model weights to disk" + color.END)

# Save the model in h5 format 
model.save("finalized_keras_model.h5")
print(color.GREEN + "Saved model to disk" +  color.END)

Saved model weights to disk
Saved model to disk

Design, train and test LSTM classifiers¶

Model with Text Input¶

# Select input and output features
industry_df['input_desc'] = industry_df['cleaned_description'] + " " +industry_df['critical_risk']
industry_df['input_desc'] = industry_df['input_desc'].str.replace('\d+', '')
X_text = industry_df['input_desc']
y_text = industry_df['accident_level']
y_text = LabelEncoder().fit_transform(y_text)

# Divide our data into testing and training sets:
X_text_train, X_text_test, y_text_train, y_text_test = train_test_split(X_text, y_text, test_size = 0.20, random_state = 1, stratify = y_text)

print(color.GREEN + 'X_text_train shape : ' + color.END + str(X_text_train.shape[0]))
print(color.GREEN + 'y_text_train shape : ' + color.END + str(y_text_train.shape[0]))
print(color.GREEN + 'X_text_test shape : ' + color.END + str(X_text_test.shape[0]))
print(color.GREEN + 'y_text_test shape : ' + color.END + str(y_text_test.shape[0]))

X_text_train shape : 334
y_text_train shape : 334
X_text_test shape : 84
y_text_test shape : 84

# Convert both the training and test labels into one-hot encoded vectors:
y_text_train = np_utils.to_categorical(y_text_train)
y_text_test = np_utils.to_categorical(y_text_test)

# The first step in word embeddings is to convert the words into their corresponding numeric indexes.
tokenizer = Tokenizer(num_words=5000)
tokenizer.fit_on_texts(X_text_train)

X_text_train = tokenizer.texts_to_sequences(X_text_train)
X_text_test = tokenizer.texts_to_sequences(X_text_test)

# Sentences can have different lengths, and therefore the sequences returned by the Tokenizer class also consist of variable lengths.
# We need to pad the our sequences using the max length.
vocab_size = len(tokenizer.word_index) + 1
print("vocab_size:", vocab_size)

maxlen = 100

X_text_train = pad_sequences(X_text_train, padding='post', maxlen=maxlen)
X_text_test = pad_sequences(X_text_test, padding='post', maxlen=maxlen)

vocab_size: 2765

# We need to load the built-in GloVe word embeddings
embedding_size = 200
embeddings_dictionary = dict()

glove_file = open('glove.6B.200d.txt', encoding="utf8")

for line in glove_file:
    records = line.split()
    word = records[0]
    vector_dimensions = np.asarray(records[1:], dtype='float32')
    embeddings_dictionary[word] = vector_dimensions

glove_file.close()

embedding_matrix = np.zeros((vocab_size, embedding_size))

for word, index in tokenizer.word_index.items():
    embedding_vector = embeddings_dictionary.get(word)
    if embedding_vector is not None:
        embedding_matrix[index] = embedding_vector

len(embeddings_dictionary.values())

400000

random_state = 42

# Build a LSTM Neural Network
deep_inputs = Input(shape=(maxlen,))
embedding_layer = Embedding(vocab_size, embedding_size, weights=[embedding_matrix], trainable=False)(deep_inputs)

LSTM_Layer_1 = Bidirectional(LSTM(128, return_sequences = True))(embedding_layer)
max_pool_layer_1 = GlobalMaxPool1D()(LSTM_Layer_1)
drop_out_layer_1 = Dropout(0.5, input_shape = (256,))(max_pool_layer_1)
dense_layer_1 = Dense(128, activation = 'relu')(drop_out_layer_1)
drop_out_layer_2 = Dropout(0.5, input_shape = (128,))(dense_layer_1)
dense_layer_2 = Dense(64, activation = 'relu')(drop_out_layer_2)
drop_out_layer_3 = Dropout(0.5, input_shape = (64,))(dense_layer_2)

dense_layer_3 = Dense(32, activation = 'relu')(drop_out_layer_3)
drop_out_layer_4 = Dropout(0.5, input_shape = (32,))(dense_layer_3)

dense_layer_4 = Dense(10, activation = 'relu')(drop_out_layer_4)
drop_out_layer_5 = Dropout(0.5, input_shape = (10,))(dense_layer_4)

dense_layer_5 = Dense(5, activation='softmax')(drop_out_layer_5)

model = Model(inputs=deep_inputs, outputs=dense_layer_5)

opt = SGD(lr=0.001, momentum=0.9)
model.compile(loss='categorical_crossentropy', optimizer=opt, metrics=['acc'])

print(model.summary())

Model: "model"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_1 (InputLayer)         [(None, 100)]             0         
_________________________________________________________________
embedding (Embedding)        (None, 100, 200)          553000    
_________________________________________________________________
bidirectional (Bidirectional (None, 100, 256)          336896    
_________________________________________________________________
global_max_pooling1d (Global (None, 256)               0         
_________________________________________________________________
dropout_4 (Dropout)          (None, 256)               0         
_________________________________________________________________
dense_11 (Dense)             (None, 128)               32896     
_________________________________________________________________
dropout_5 (Dropout)          (None, 128)               0         
_________________________________________________________________
dense_12 (Dense)             (None, 64)                8256      
_________________________________________________________________
dropout_6 (Dropout)          (None, 64)                0         
_________________________________________________________________
dense_13 (Dense)             (None, 32)                2080      
_________________________________________________________________
dropout_7 (Dropout)          (None, 32)                0         
_________________________________________________________________
dense_14 (Dense)             (None, 10)                330       
_________________________________________________________________
dropout_8 (Dropout)          (None, 10)                0         
_________________________________________________________________
dense_15 (Dense)             (None, 5)                 55        
=================================================================
Total params: 933,513
Trainable params: 380,513
Non-trainable params: 553,000
_________________________________________________________________
None

# Use earlystopping
callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=7, min_delta=1E-3)
rlrp = ReduceLROnPlateau(monitor='val_loss', factor=0.0001, patience=5, min_delta=1E-4)

target_type = 'multi_label'
metrics = Metrics(validation_data=(X_text_train, y_text_train, target_type))

# fit the keras model on the dataset
training_history = model.fit(X_text_train, y_text_train, epochs=100, batch_size=8, verbose=1, validation_data=(X_text_test, y_text_test), callbacks=[rlrp, metrics])

Epoch 1/100
42/42 [==============================] - 13s 169ms/step - loss: 1.6252 - acc: 0.3073 - val_loss: 1.3477 - val_acc: 0.7381
Epoch 2/100
42/42 [==============================] - 5s 114ms/step - loss: 1.3137 - acc: 0.6628 - val_loss: 1.1741 - val_acc: 0.7381
Epoch 3/100
42/42 [==============================] - 4s 101ms/step - loss: 1.1600 - acc: 0.7309 - val_loss: 1.0592 - val_acc: 0.7381
Epoch 4/100
42/42 [==============================] - 4s 104ms/step - loss: 1.3410 - acc: 0.7034 - val_loss: 1.0769 - val_acc: 0.7381ss: 1.4124 -  - ETA: 0s - loss: 1.3658 - acc:
Epoch 5/100
42/42 [==============================] - 5s 114ms/step - loss: 1.0966 - acc: 0.7396 - val_loss: 1.0146 - val_acc: 0.7381
Epoch 6/100
42/42 [==============================] - 5s 110ms/step - loss: 1.1030 - acc: 0.7386 - val_loss: 1.0176 - val_acc: 0.7381
Epoch 7/100
42/42 [==============================] - 5s 116ms/step - loss: 1.0122 - acc: 0.7817 - val_loss: 1.0033 - val_acc: 0.7381
Epoch 8/100
42/42 [==============================] - 5s 108ms/step - loss: 1.0613 - acc: 0.7423 - val_loss: 0.9924 - val_acc: 0.7381
Epoch 9/100
42/42 [==============================] - 4s 99ms/step - loss: 0.9996 - acc: 0.7416 - val_loss: 0.9721 - val_acc: 0.7381
Epoch 10/100
42/42 [==============================] - 4s 104ms/step - loss: 0.9469 - acc: 0.7867 - val_loss: 0.9615 - val_acc: 0.7381
Epoch 11/100
42/42 [==============================] - 5s 110ms/step - loss: 1.0368 - acc: 0.7192 - val_loss: 0.9539 - val_acc: 0.7381
Epoch 12/100
42/42 [==============================] - 4s 103ms/step - loss: 0.8874 - acc: 0.7695 - val_loss: 0.9472 - val_acc: 0.7381
Epoch 13/100
42/42 [==============================] - 4s 99ms/step - loss: 0.9413 - acc: 0.7617 - val_loss: 0.9439 - val_acc: 0.7381
Epoch 14/100
42/42 [==============================] - 4s 101ms/step - loss: 0.9411 - acc: 0.7502 - val_loss: 0.9392 - val_acc: 0.7381: 0s - loss: 0.9367 - acc: 0. - ETA: 0s - loss: 0.9395 - acc: 0.75
Epoch 15/100
42/42 [==============================] - 4s 102ms/step - loss: 0.9695 - acc: 0.7488 - val_loss: 0.9400 - val_acc: 0.7381
Epoch 16/100
42/42 [==============================] - 4s 94ms/step - loss: 0.8907 - acc: 0.7615 - val_loss: 0.9370 - val_acc: 0.7381
Epoch 17/100
42/42 [==============================] - 4s 104ms/step - loss: 0.9793 - acc: 0.7338 - val_loss: 0.9358 - val_acc: 0.7381
Epoch 18/100
42/42 [==============================] - 4s 103ms/step - loss: 0.9378 - acc: 0.7345 - val_loss: 0.9310 - val_acc: 0.7381
Epoch 19/100
42/42 [==============================] - 4s 101ms/step - loss: 0.8333 - acc: 0.7851 - val_loss: 0.9272 - val_acc: 0.7381
Epoch 20/100
42/42 [==============================] - 6s 142ms/step - loss: 1.0314 - acc: 0.7210 - val_loss: 0.9299 - val_acc: 0.7381
Epoch 21/100
42/42 [==============================] - 4s 100ms/step - loss: 0.9011 - acc: 0.7669 - val_loss: 0.9300 - val_acc: 0.7381
Epoch 22/100
42/42 [==============================] - 5s 109ms/step - loss: 1.0148 - acc: 0.6944 - val_loss: 0.9296 - val_acc: 0.7381
Epoch 23/100
42/42 [==============================] - 5s 113ms/step - loss: 1.0040 - acc: 0.7093 - val_loss: 0.9269 - val_acc: 0.7381
Epoch 24/100
42/42 [==============================] - 4s 104ms/step - loss: 0.8730 - acc: 0.7660 - val_loss: 0.9238 - val_acc: 0.7381
Epoch 25/100
42/42 [==============================] - 5s 108ms/step - loss: 0.8796 - acc: 0.7509 - val_loss: 0.9237 - val_acc: 0.7381
Epoch 26/100
42/42 [==============================] - 4s 99ms/step - loss: 0.8663 - acc: 0.7719 - val_loss: 0.9234 - val_acc: 0.7381
Epoch 27/100
42/42 [==============================] - 4s 101ms/step - loss: 0.9645 - acc: 0.7271 - val_loss: 0.9226 - val_acc: 0.7381
Epoch 28/100
42/42 [==============================] - 5s 113ms/step - loss: 0.8730 - acc: 0.7628 - val_loss: 0.9211 - val_acc: 0.7381
Epoch 29/100
42/42 [==============================] - 5s 116ms/step - loss: 0.9229 - acc: 0.7406 - val_loss: 0.9224 - val_acc: 0.7381
Epoch 30/100
42/42 [==============================] - 5s 114ms/step - loss: 0.8810 - acc: 0.7652 - val_loss: 0.9234 - val_acc: 0.7381
Epoch 31/100
42/42 [==============================] - 4s 102ms/step - loss: 0.9339 - acc: 0.7407 - val_loss: 0.9226 - val_acc: 0.7381
Epoch 32/100
42/42 [==============================] - 4s 101ms/step - loss: 0.9015 - acc: 0.7582 - val_loss: 0.9210 - val_acc: 0.7381
Epoch 33/100
42/42 [==============================] - 4s 94ms/step - loss: 0.8587 - acc: 0.7760 - val_loss: 0.9210 - val_acc: 0.7381
Epoch 34/100
42/42 [==============================] - 5s 118ms/step - loss: 0.9590 - acc: 0.7256 - val_loss: 0.9209 - val_acc: 0.7381
Epoch 35/100
42/42 [==============================] - 4s 103ms/step - loss: 0.8589 - acc: 0.7737 - val_loss: 0.9202 - val_acc: 0.7381
Epoch 36/100
42/42 [==============================] - 4s 97ms/step - loss: 0.8904 - acc: 0.7470 - val_loss: 0.9201 - val_acc: 0.7381
Epoch 37/100
42/42 [==============================] - 4s 94ms/step - loss: 0.9504 - acc: 0.7260 - val_loss: 0.9198 - val_acc: 0.7381
Epoch 38/100
42/42 [==============================] - 4s 101ms/step - loss: 0.9369 - acc: 0.7300 - val_loss: 0.9193 - val_acc: 0.7381
Epoch 39/100
42/42 [==============================] - 5s 107ms/step - loss: 0.8929 - acc: 0.7425 - val_loss: 0.9180 - val_acc: 0.7381
Epoch 40/100
42/42 [==============================] - 5s 119ms/step - loss: 0.9133 - acc: 0.7408 - val_loss: 0.9180 - val_acc: 0.7381
Epoch 41/100
42/42 [==============================] - 5s 107ms/step - loss: 0.9539 - acc: 0.7275 - val_loss: 0.9181 - val_acc: 0.7381
Epoch 42/100
42/42 [==============================] - 5s 113ms/step - loss: 1.0063 - acc: 0.7093 - val_loss: 0.9179 - val_acc: 0.7381
Epoch 43/100
42/42 [==============================] - 5s 108ms/step - loss: 0.8579 - acc: 0.7632 - val_loss: 0.9177 - val_acc: 0.7381
Epoch 44/100
42/42 [==============================] - 5s 111ms/step - loss: 0.8754 - acc: 0.7625 - val_loss: 0.9174 - val_acc: 0.7381
Epoch 45/100
42/42 [==============================] - 5s 111ms/step - loss: 0.9191 - acc: 0.7388 - val_loss: 0.9174 - val_acc: 0.7381
Epoch 46/100
42/42 [==============================] - 5s 111ms/step - loss: 0.8956 - acc: 0.7490 - val_loss: 0.9172 - val_acc: 0.7381
Epoch 47/100
42/42 [==============================] - 5s 119ms/step - loss: 1.0534 - acc: 0.6878 - val_loss: 0.9172 - val_acc: 0.7381
Epoch 48/100
42/42 [==============================] - 5s 109ms/step - loss: 0.9234 - acc: 0.7411 - val_loss: 0.9165 - val_acc: 0.7381
Epoch 49/100
42/42 [==============================] - 4s 103ms/step - loss: 0.8918 - acc: 0.7465 - val_loss: 0.9162 - val_acc: 0.7381
Epoch 50/100
42/42 [==============================] - 5s 108ms/step - loss: 0.9468 - acc: 0.7343 - val_loss: 0.9159 - val_acc: 0.7381
Epoch 51/100
42/42 [==============================] - 4s 103ms/step - loss: 0.9837 - acc: 0.7049 - val_loss: 0.9157 - val_acc: 0.7381
Epoch 52/100
42/42 [==============================] - 5s 116ms/step - loss: 0.9067 - acc: 0.7419 - val_loss: 0.9151 - val_acc: 0.7381
Epoch 53/100
42/42 [==============================] - 5s 125ms/step - loss: 0.9252 - acc: 0.7465 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 54/100
42/42 [==============================] - 5s 123ms/step - loss: 0.9736 - acc: 0.7156 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 55/100
42/42 [==============================] - 5s 116ms/step - loss: 0.8984 - acc: 0.7365 - val_loss: 0.9152 - val_acc: 0.7381
Epoch 56/100
42/42 [==============================] - 5s 113ms/step - loss: 0.8594 - acc: 0.7642 - val_loss: 0.9152 - val_acc: 0.7381
Epoch 57/100
42/42 [==============================] - 5s 111ms/step - loss: 0.8954 - acc: 0.7513 - val_loss: 0.9155 - val_acc: 0.7381
Epoch 58/100
42/42 [==============================] - 4s 96ms/step - loss: 0.9663 - acc: 0.7199 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 59/100
42/42 [==============================] - 5s 118ms/step - loss: 0.8586 - acc: 0.7640 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 60/100
42/42 [==============================] - 4s 105ms/step - loss: 0.9534 - acc: 0.7187 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 61/100
42/42 [==============================] - 5s 110ms/step - loss: 0.9740 - acc: 0.7131 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 62/100
42/42 [==============================] - 4s 102ms/step - loss: 0.9324 - acc: 0.7353 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 63/100
42/42 [==============================] - 5s 111ms/step - loss: 0.8751 - acc: 0.7597 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 64/100
42/42 [==============================] - 4s 103ms/step - loss: 0.9706 - acc: 0.7075 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 65/100
42/42 [==============================] - 5s 112ms/step - loss: 0.9523 - acc: 0.7116 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 66/100
42/42 [==============================] - 5s 112ms/step - loss: 0.9454 - acc: 0.7220 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 67/100
42/42 [==============================] - 4s 100ms/step - loss: 0.8985 - acc: 0.7422 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 68/100
42/42 [==============================] - 5s 109ms/step - loss: 0.9846 - acc: 0.7083 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 69/100
42/42 [==============================] - 4s 104ms/step - loss: 0.9306 - acc: 0.7367 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 70/100
42/42 [==============================] - 5s 113ms/step - loss: 0.8776 - acc: 0.7496 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 71/100
42/42 [==============================] - 5s 125ms/step - loss: 0.9505 - acc: 0.7223 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 72/100
42/42 [==============================] - 4s 106ms/step - loss: 0.9880 - acc: 0.7138 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 73/100
42/42 [==============================] - 6s 135ms/step - loss: 0.8428 - acc: 0.7683 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 74/100
42/42 [==============================] - 5s 124ms/step - loss: 0.8696 - acc: 0.7555 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 75/100
42/42 [==============================] - 5s 124ms/step - loss: 0.8826 - acc: 0.7509 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 76/100
42/42 [==============================] - 5s 130ms/step - loss: 0.9051 - acc: 0.7371 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 77/100
42/42 [==============================] - 5s 120ms/step - loss: 0.9089 - acc: 0.7435 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 78/100
42/42 [==============================] - 5s 121ms/step - loss: 0.8958 - acc: 0.7517 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 79/100
42/42 [==============================] - 5s 122ms/step - loss: 0.8768 - acc: 0.7537 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 80/100
42/42 [==============================] - 5s 121ms/step - loss: 0.9242 - acc: 0.7305 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 81/100
42/42 [==============================] - 4s 106ms/step - loss: 0.8946 - acc: 0.7464 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 82/100
42/42 [==============================] - 5s 116ms/step - loss: 0.9114 - acc: 0.7456 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 83/100
42/42 [==============================] - 5s 120ms/step - loss: 0.9188 - acc: 0.7247 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 84/100
42/42 [==============================] - 5s 124ms/step - loss: 0.8228 - acc: 0.7740 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 85/100
42/42 [==============================] - 5s 121ms/step - loss: 0.9020 - acc: 0.7485 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 86/100
42/42 [==============================] - 5s 120ms/step - loss: 0.8927 - acc: 0.7424 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 87/100
42/42 [==============================] - 5s 120ms/step - loss: 0.9072 - acc: 0.7330 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 88/100
42/42 [==============================] - 4s 108ms/step - loss: 0.9204 - acc: 0.7352 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 89/100
42/42 [==============================] - 4s 107ms/step - loss: 0.8327 - acc: 0.7732 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 90/100
42/42 [==============================] - 4s 106ms/step - loss: 0.9293 - acc: 0.7358 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 91/100
42/42 [==============================] - 5s 114ms/step - loss: 0.8981 - acc: 0.7492 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 92/100
42/42 [==============================] - 4s 105ms/step - loss: 0.9202 - acc: 0.7384 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 93/100
42/42 [==============================] - 5s 117ms/step - loss: 0.8705 - acc: 0.7564 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 94/100
42/42 [==============================] - 4s 106ms/step - loss: 0.9249 - acc: 0.7382 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 95/100
42/42 [==============================] - 4s 107ms/step - loss: 0.9413 - acc: 0.7271 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 96/100
42/42 [==============================] - 4s 106ms/step - loss: 0.8031 - acc: 0.7866 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 97/100
42/42 [==============================] - 5s 115ms/step - loss: 0.9618 - acc: 0.7155 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 98/100
42/42 [==============================] - 5s 111ms/step - loss: 0.9482 - acc: 0.7255 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 99/100
42/42 [==============================] - 5s 111ms/step - loss: 0.8890 - acc: 0.7504 - val_loss: 0.9156 - val_acc: 0.7381
Epoch 100/100
42/42 [==============================] - 5s 115ms/step - loss: 0.9717 - acc: 0.7125 - val_loss: 0.9156 - val_acc: 0.7381

# evaluate the keras model
_, train_accuracy = model.evaluate(X_text_train, y_text_train, batch_size=8, verbose=0)
_, test_accuracy = model.evaluate(X_text_test, y_text_test, batch_size=8, verbose=0)

print(color.GREEN + 'Train accuracy: ' + color.END + str(train_accuracy*100))
print(color.GREEN + 'Test accuracy: ' + color.END + str(test_accuracy*100))

Train accuracy: 73.95209670066833
Test accuracy: 73.8095223903656

accuracy, precision, recall, f1 = get_classification_metrics(model, X_text_test, y_text_test, target_type)
print(color.BLUE + 'Accuracy: ' + color.END + str(accuracy))
print(color.BLUE + 'Precision: ' + color.END + str(precision))
print(color.BLUE + 'Recall: ' + color.END + str(recall))
print(color.BLUE + 'F1 score:' + color.END + str(f1))

Accuracy: 0.7380952380952381
Precision: 0.7380952380952381
Recall: 0.7380952380952381
F1 score:0.7380952380952381

epochs = range(len(training_history.history['loss'])) # Get number of epochs

# plot loss learning curves
plt.plot(epochs, training_history.history['loss'], label = 'train')
plt.plot(epochs, training_history.history['val_loss'], label = 'test')
plt.legend(loc = 'upper right')
plt.title ('Training and validation loss')

Text(0.5,1,'Training and validation loss')

# plot accuracy learning curves
plt.plot(epochs, training_history.history['acc'], label = 'train')
plt.plot(epochs, training_history.history['val_acc'], label = 'test')
plt.legend(loc = 'upper right')
plt.title ('Training and validation accuracy')

Text(0.5,1,'Training and validation accuracy')

# serialize model to JSON
model_json = model.to_json()
with open("model.json", "w") as json_file:
    json_file.write(model_json)
    
# serialize weights to HDF5
model.save_weights("model.h5")
print("Saved model weights to disk")

# Save the model in h5 format 
model.save("finalized_keras_model_v0.h5")
print("Saved model to disk")

Saved model weights to disk
Saved model to disk

Model using Categorical Values¶

# Select input and output features
X_cat = industry_featenc_df.drop(['accident_level','potential_accident_level'], axis = 1)
y_cat = industry_df['accident_level']

y_cat = LabelEncoder().fit_transform(y_cat)

X_cat_train, X_cat_test, y_cat_train, y_cat_test = train_test_split(X_cat, y_cat, test_size = 0.20, random_state = 1, stratify = y_cat)

print(color.GREEN + 'X_cat_train shape : ' + color.END + str(X_cat_train.shape[0]))
print(color.GREEN + 'y_cat_train shape : ' + color.END + str(y_cat_train.shape[0]))
print(color.GREEN + 'X_cat_test shape : ' + color.END + str(X_cat_test.shape[0]))
print(color.GREEN + 'y_cat_test shape : ' + color.END + str(y_cat_test.shape[0]))

X_cat_train shape : 334
y_cat_train shape : 334
X_cat_test shape : 84
y_cat_test shape : 84

y_cat_train = np_utils.to_categorical(y_cat_train)
y_cat_test = np_utils.to_categorical(y_cat_test)

# Variable transformation using StandardScaler
scaler_X = StandardScaler()#StandardScaler()
X_cat_train.iloc[:,:12] = scaler_X.fit_transform(X_cat_train.iloc[:,:12]) # Scaling only first 6 feautres

X_cat_test.iloc[:,:12] = scaler_X.fit_transform(X_cat_test.iloc[:,:12]) # Scaling only first 6 feautres

param = 1e-4

input2 = Input(shape=(X_cat_train.shape[1],))
dense_layer_1 = Dense(10, input_dim=X_cat_train.shape[1], activation='relu', kernel_initializer='he_uniform', kernel_regularizer=l2(param),
                kernel_constraint=unit_norm())(input2)
drop_out_layer_1 = Dropout(0.2)(dense_layer_1)
batch_norm_layer_1 = BatchNormalization()(drop_out_layer_1)
dense_layer_2 = Dense(10, activation='relu', kernel_initializer='he_uniform', kernel_regularizer=l2(param), 
                kernel_constraint=unit_norm())(batch_norm_layer_1)
drop_out_layer_2 = Dropout(0.5)(dense_layer_2)
batch_norm_layer_2 = BatchNormalization()(drop_out_layer_2)
dense_layer_3 = Dense(5, activation='softmax', kernel_regularizer=l2(param), kernel_constraint=unit_norm())(batch_norm_layer_2)

model = Model(inputs=input2, outputs=dense_layer_3)

# compile the keras model
#opt = optimizers.Adamax(lr=0.01)
opt = SGD(lr=0.001, momentum=0.9)
model.compile(loss='categorical_crossentropy', optimizer=opt, metrics=['acc'])

print(model.summary())

Model: "model_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_2 (InputLayer)         [(None, 12)]              0         
_________________________________________________________________
dense_16 (Dense)             (None, 10)                130       
_________________________________________________________________
dropout_9 (Dropout)          (None, 10)                0         
_________________________________________________________________
batch_normalization_4 (Batch (None, 10)                40        
_________________________________________________________________
dense_17 (Dense)             (None, 10)                110       
_________________________________________________________________
dropout_10 (Dropout)         (None, 10)                0         
_________________________________________________________________
batch_normalization_5 (Batch (None, 10)                40        
_________________________________________________________________
dense_18 (Dense)             (None, 5)                 55        
=================================================================
Total params: 375
Trainable params: 335
Non-trainable params: 40
_________________________________________________________________
None

# Use earlystopping
callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=7, min_delta=1E-3)
rlrp = ReduceLROnPlateau(monitor='val_loss', factor=0.0001, patience=5, min_delta=1E-4)

target_type = 'multi_label'
metrics = Metrics(validation_data=(X_cat_train, y_cat_train, target_type))

# fit the keras model on the dataset
training_history = model.fit(X_cat_train, y_cat_train, epochs=100, batch_size=8, verbose=1, validation_data=(X_cat_test, y_cat_test), callbacks=[rlrp, metrics])

Epoch 1/100
42/42 [==============================] - 2s 15ms/step - loss: 1.9775 - acc: 0.2394 - val_loss: 1.3390 - val_acc: 0.5595
Epoch 2/100
42/42 [==============================] - 0s 5ms/step - loss: 1.6395 - acc: 0.4161 - val_loss: 1.2014 - val_acc: 0.6667
Epoch 3/100
42/42 [==============================] - 0s 7ms/step - loss: 1.5061 - acc: 0.4335 - val_loss: 1.1241 - val_acc: 0.7143
Epoch 4/100
42/42 [==============================] - 0s 5ms/step - loss: 1.3309 - acc: 0.5244 - val_loss: 1.0790 - val_acc: 0.7381
Epoch 5/100
42/42 [==============================] - 0s 5ms/step - loss: 1.2454 - acc: 0.6239 - val_loss: 1.0538 - val_acc: 0.7381
Epoch 6/100
42/42 [==============================] - 0s 5ms/step - loss: 1.0485 - acc: 0.6944 - val_loss: 1.0246 - val_acc: 0.7381
Epoch 7/100
42/42 [==============================] - 0s 12ms/step - loss: 0.9704 - acc: 0.7513 - val_loss: 1.0212 - val_acc: 0.7381
Epoch 8/100
42/42 [==============================] - 0s 5ms/step - loss: 1.0776 - acc: 0.7038 - val_loss: 1.0133 - val_acc: 0.7381
Epoch 9/100
42/42 [==============================] - 0s 9ms/step - loss: 0.9728 - acc: 0.7226 - val_loss: 1.0007 - val_acc: 0.7381
Epoch 10/100
42/42 [==============================] - 0s 6ms/step - loss: 0.8858 - acc: 0.7675 - val_loss: 0.9926 - val_acc: 0.7381
Epoch 11/100
42/42 [==============================] - 0s 5ms/step - loss: 1.0791 - acc: 0.6921 - val_loss: 0.9885 - val_acc: 0.7381
Epoch 12/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9816 - acc: 0.7522 - val_loss: 0.9794 - val_acc: 0.7381
Epoch 13/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9573 - acc: 0.7558 - val_loss: 0.9758 - val_acc: 0.7381
Epoch 14/100
42/42 [==============================] - 0s 6ms/step - loss: 0.9588 - acc: 0.7502 - val_loss: 0.9669 - val_acc: 0.7381
Epoch 15/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9438 - acc: 0.7488 - val_loss: 0.9632 - val_acc: 0.7381
Epoch 16/100
42/42 [==============================] - 0s 8ms/step - loss: 0.8917 - acc: 0.7579 - val_loss: 0.9580 - val_acc: 0.7381
Epoch 17/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9812 - acc: 0.7329 - val_loss: 0.9559 - val_acc: 0.7381
Epoch 18/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9953 - acc: 0.7301 - val_loss: 0.9575 - val_acc: 0.7381
Epoch 19/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8779 - acc: 0.7851 - val_loss: 0.9572 - val_acc: 0.7381
Epoch 20/100
42/42 [==============================] - 0s 10ms/step - loss: 1.0292 - acc: 0.7210 - val_loss: 0.9580 - val_acc: 0.7381
Epoch 21/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8772 - acc: 0.7586 - val_loss: 0.9565 - val_acc: 0.7381
Epoch 22/100
42/42 [==============================] - ETA: 0s - loss: 1.1059 - acc: 0.675 - 0s 6ms/step - loss: 1.0622 - acc: 0.6944 - val_loss: 0.9516 - val_acc: 0.7381
Epoch 23/100
42/42 [==============================] - 0s 6ms/step - loss: 1.0351 - acc: 0.7093 - val_loss: 0.9485 - val_acc: 0.7381
Epoch 24/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8697 - acc: 0.7660 - val_loss: 0.9471 - val_acc: 0.7381
Epoch 25/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9046 - acc: 0.7494 - val_loss: 0.9428 - val_acc: 0.7381
Epoch 26/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8771 - acc: 0.7719 - val_loss: 0.9443 - val_acc: 0.7381
Epoch 27/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9602 - acc: 0.7254 - val_loss: 0.9432 - val_acc: 0.7381
Epoch 28/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8785 - acc: 0.7628 - val_loss: 0.9417 - val_acc: 0.7381
Epoch 29/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9615 - acc: 0.7406 - val_loss: 0.9421 - val_acc: 0.7381
Epoch 30/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8626 - acc: 0.7652 - val_loss: 0.9409 - val_acc: 0.7381
Epoch 31/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9386 - acc: 0.7407 - val_loss: 0.9394 - val_acc: 0.7381
Epoch 32/100
42/42 [==============================] - 0s 6ms/step - loss: 0.8752 - acc: 0.7582 - val_loss: 0.9371 - val_acc: 0.7381
Epoch 33/100
42/42 [==============================] - 0s 11ms/step - loss: 0.8413 - acc: 0.7760 - val_loss: 0.9366 - val_acc: 0.7381
Epoch 34/100
42/42 [==============================] - 0s 6ms/step - loss: 0.9642 - acc: 0.7256 - val_loss: 0.9371 - val_acc: 0.7381
Epoch 35/100
42/42 [==============================] - 0s 6ms/step - loss: 0.8492 - acc: 0.7737 - val_loss: 0.9363 - val_acc: 0.7381
Epoch 36/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8899 - acc: 0.7470 - val_loss: 0.9346 - val_acc: 0.7381
Epoch 37/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9444 - acc: 0.7260 - val_loss: 0.9341 - val_acc: 0.7381
Epoch 38/100
42/42 [==============================] - 0s 6ms/step - loss: 0.9343 - acc: 0.7300 - val_loss: 0.9336 - val_acc: 0.7381
Epoch 39/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9151 - acc: 0.7425 - val_loss: 0.9324 - val_acc: 0.7381
Epoch 40/100
42/42 [==============================] - 0s 7ms/step - loss: 0.9438 - acc: 0.7408 - val_loss: 0.9321 - val_acc: 0.7381
Epoch 41/100
42/42 [==============================] - 0s 7ms/step - loss: 0.9579 - acc: 0.7275 - val_loss: 0.9318 - val_acc: 0.7381
Epoch 42/100
42/42 [==============================] - 0s 10ms/step - loss: 0.9888 - acc: 0.7093 - val_loss: 0.9299 - val_acc: 0.7381
Epoch 43/100
42/42 [==============================] - 1s 12ms/step - loss: 0.8533 - acc: 0.7632 - val_loss: 0.9292 - val_acc: 0.7381
Epoch 44/100
42/42 [==============================] - 0s 7ms/step - loss: 0.8566 - acc: 0.7625 - val_loss: 0.9274 - val_acc: 0.7381
Epoch 45/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9330 - acc: 0.7388 - val_loss: 0.9273 - val_acc: 0.7381
Epoch 46/100
42/42 [==============================] - 0s 6ms/step - loss: 0.9087 - acc: 0.7490 - val_loss: 0.9268 - val_acc: 0.7381
Epoch 47/100
42/42 [==============================] - 1s 13ms/step - loss: 1.0473 - acc: 0.6878 - val_loss: 0.9237 - val_acc: 0.7381
Epoch 48/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9118 - acc: 0.7411 - val_loss: 0.9226 - val_acc: 0.7381
Epoch 49/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8792 - acc: 0.7465 - val_loss: 0.9221 - val_acc: 0.7381
Epoch 50/100
42/42 [==============================] - 0s 6ms/step - loss: 0.9233 - acc: 0.7343 - val_loss: 0.9228 - val_acc: 0.7381
Epoch 51/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9935 - acc: 0.7049 - val_loss: 0.9242 - val_acc: 0.7381
Epoch 52/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8873 - acc: 0.7419 - val_loss: 0.9231 - val_acc: 0.7381
Epoch 53/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8975 - acc: 0.7465 - val_loss: 0.9246 - val_acc: 0.7381
Epoch 54/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9713 - acc: 0.7156 - val_loss: 0.9220 - val_acc: 0.7381
Epoch 55/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9264 - acc: 0.7365 - val_loss: 0.9218 - val_acc: 0.7381
Epoch 56/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8657 - acc: 0.7642 - val_loss: 0.9218 - val_acc: 0.7381
Epoch 57/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9034 - acc: 0.7513 - val_loss: 0.9222 - val_acc: 0.7381
Epoch 58/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9484 - acc: 0.7199 - val_loss: 0.9230 - val_acc: 0.7381
Epoch 59/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8418 - acc: 0.7640 - val_loss: 0.9229 - val_acc: 0.7381
Epoch 60/100
42/42 [==============================] - 0s 4ms/step - loss: 0.9382 - acc: 0.7187 - val_loss: 0.9234 - val_acc: 0.7381
Epoch 61/100
42/42 [==============================] - 0s 11ms/step - loss: 0.9773 - acc: 0.7131 - val_loss: 0.9232 - val_acc: 0.7381
Epoch 62/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9360 - acc: 0.7353 - val_loss: 0.9228 - val_acc: 0.7381
Epoch 63/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8514 - acc: 0.7597 - val_loss: 0.9230 - val_acc: 0.7381
Epoch 64/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9842 - acc: 0.7075 - val_loss: 0.9239 - val_acc: 0.7381
Epoch 65/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9979 - acc: 0.7116 - val_loss: 0.9234 - val_acc: 0.7381
Epoch 66/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9431 - acc: 0.7220 - val_loss: 0.9236 - val_acc: 0.7381
Epoch 67/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9097 - acc: 0.7422 - val_loss: 0.9220 - val_acc: 0.7381
Epoch 68/100
42/42 [==============================] - 0s 6ms/step - loss: 0.9947 - acc: 0.7083 - val_loss: 0.9225 - val_acc: 0.7381
Epoch 69/100
42/42 [==============================] - 0s 6ms/step - loss: 0.8911 - acc: 0.7367 - val_loss: 0.9225 - val_acc: 0.7381
Epoch 70/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8863 - acc: 0.7496 - val_loss: 0.9224 - val_acc: 0.7381
Epoch 71/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9440 - acc: 0.7223 - val_loss: 0.9220 - val_acc: 0.7381
Epoch 72/100
42/42 [==============================] - 0s 6ms/step - loss: 1.0033 - acc: 0.7138 - val_loss: 0.9215 - val_acc: 0.7381
Epoch 73/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8385 - acc: 0.7683 - val_loss: 0.9215 - val_acc: 0.7381
Epoch 74/100
42/42 [==============================] - 0s 10ms/step - loss: 0.8517 - acc: 0.7555 - val_loss: 0.9215 - val_acc: 0.7381
Epoch 75/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8500 - acc: 0.7509 - val_loss: 0.9217 - val_acc: 0.7381
Epoch 76/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8855 - acc: 0.7371 - val_loss: 0.9215 - val_acc: 0.7381
Epoch 77/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8885 - acc: 0.7435 - val_loss: 0.9214 - val_acc: 0.7381
Epoch 78/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8920 - acc: 0.7517 - val_loss: 0.9219 - val_acc: 0.7381
Epoch 79/100
42/42 [==============================] - 0s 7ms/step - loss: 0.8730 - acc: 0.7537 - val_loss: 0.9222 - val_acc: 0.7381
Epoch 80/100
42/42 [==============================] - 0s 8ms/step - loss: 0.9142 - acc: 0.7305 - val_loss: 0.9220 - val_acc: 0.7381
Epoch 81/100
42/42 [==============================] - 0s 7ms/step - loss: 0.9082 - acc: 0.7464 - val_loss: 0.9222 - val_acc: 0.7381
Epoch 82/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8887 - acc: 0.7456 - val_loss: 0.9208 - val_acc: 0.7381
Epoch 83/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9079 - acc: 0.7247 - val_loss: 0.9209 - val_acc: 0.7381
Epoch 84/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8075 - acc: 0.7740 - val_loss: 0.9216 - val_acc: 0.7381
Epoch 85/100
42/42 [==============================] - 0s 10ms/step - loss: 0.8551 - acc: 0.7485 - val_loss: 0.9206 - val_acc: 0.7381
Epoch 86/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8617 - acc: 0.7424 - val_loss: 0.9206 - val_acc: 0.7381
Epoch 87/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9254 - acc: 0.7330 - val_loss: 0.9206 - val_acc: 0.7381
Epoch 88/100
42/42 [==============================] - 1s 14ms/step - loss: 0.8970 - acc: 0.7352 - val_loss: 0.9220 - val_acc: 0.7381
Epoch 89/100
42/42 [==============================] - 0s 5ms/step - loss: 0.7990 - acc: 0.7732 - val_loss: 0.9218 - val_acc: 0.7381
Epoch 90/100
42/42 [==============================] - 0s 6ms/step - loss: 0.9150 - acc: 0.7358 - val_loss: 0.9213 - val_acc: 0.7381
Epoch 91/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8959 - acc: 0.7492 - val_loss: 0.9217 - val_acc: 0.7381
Epoch 92/100
42/42 [==============================] - 0s 6ms/step - loss: 0.9392 - acc: 0.7384 - val_loss: 0.9213 - val_acc: 0.7381
Epoch 93/100
42/42 [==============================] - 0s 5ms/step - loss: 0.8481 - acc: 0.7564 - val_loss: 0.9216 - val_acc: 0.7381
Epoch 94/100
42/42 [==============================] - 0s 6ms/step - loss: 0.8854 - acc: 0.7382 - val_loss: 0.9212 - val_acc: 0.7381
Epoch 95/100
42/42 [==============================] - 0s 6ms/step - loss: 0.9395 - acc: 0.7271 - val_loss: 0.9218 - val_acc: 0.7381
Epoch 96/100
42/42 [==============================] - 0s 6ms/step - loss: 0.8082 - acc: 0.7866 - val_loss: 0.9220 - val_acc: 0.7381
Epoch 97/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9642 - acc: 0.7155 - val_loss: 0.9217 - val_acc: 0.7381
Epoch 98/100
42/42 [==============================] - 0s 5ms/step - loss: 0.9871 - acc: 0.7255 - val_loss: 0.9215 - val_acc: 0.7381
Epoch 99/100
42/42 [==============================] - 0s 6ms/step - loss: 0.8796 - acc: 0.7504 - val_loss: 0.9214 - val_acc: 0.7381
Epoch 100/100
42/42 [==============================] - 0s 6ms/step - loss: 0.9418 - acc: 0.7125 - val_loss: 0.9213 - val_acc: 0.7381

# evaluate the keras model
_, train_accuracy = model.evaluate(X_cat_train, y_cat_train, batch_size=8, verbose=0)
_, test_accuracy = model.evaluate(X_cat_test, y_cat_test, batch_size=8, verbose=0)

print(color.GREEN + 'Train accuracy: ' + color.END + str(train_accuracy*100))
print(color.GREEN + 'Test accuracy: ' + color.END + str(test_accuracy*100))

Train accuracy: 73.95209670066833
Test accuracy: 73.8095223903656

accuracy, precision, recall, f1 = get_classification_metrics(model, X_cat_test, y_cat_test, target_type)
print(color.BLUE + 'Accuracy: ' + color.END + str(accuracy))
print(color.BLUE + 'Precision:' + color.END + str(precision))
print(color.BLUE + 'Recall: ' + color.END + str(recall))
print(color.BLUE + 'F1 score: ' + color.END + str(f1))

Accuracy: 0.7380952380952381
Precision:0.7380952380952381
Recall: 0.7380952380952381
F1 score: 0.7380952380952381

epochs = range(len(training_history.history['loss'])) # Get number of epochs

# plot loss learning curves
plt.plot(epochs, training_history.history['loss'], label = 'train')
plt.plot(epochs, training_history.history['val_loss'], label = 'test')
plt.legend(loc = 'upper right')
plt.title ('Training and validation loss')

Text(0.5,1,'Training and validation loss')

# plot accuracy learning curves
plt.plot(epochs, training_history.history['acc'], label = 'train')
plt.plot(epochs, training_history.history['val_acc'], label = 'test')
plt.legend(loc = 'upper right')
plt.title ('Training and validation accuracy')

Text(0.5,1,'Training and validation accuracy')

# serialize model to JSON
model_json = model.to_json()
with open("model.json", "w") as json_file:
    json_file.write(model_json)
    
# serialize weights to HDF5
model.save_weights("model.h5")
print("Saved model weights to disk")

# Save the model in h5 format 
model.save("finalized_keras_model_v1.h5")
print("Saved model to disk")

Saved model weights to disk
Saved model to disk

Model using Text Inputs and Categorical Values¶

input_1 = Input(shape=(maxlen,))
embedding_layer   = Embedding(vocab_size, embedding_size, weights=[embedding_matrix], trainable=False)(input_1)
LSTM_Layer_1      = Bidirectional(LSTM(128, return_sequences = True))(embedding_layer)
max_pool_layer_1  = GlobalMaxPool1D()(LSTM_Layer_1)
drop_out_layer_1  = Dropout(0.5, input_shape = (256,))(max_pool_layer_1)
dense_layer_1     = Dense(128, activation = 'relu')(drop_out_layer_1)
drop_out_layer_2  = Dropout(0.5, input_shape = (128,))(dense_layer_1)
dense_layer_2     = Dense(64, activation = 'relu')(drop_out_layer_2)
drop_out_layer_3  = Dropout(0.5, input_shape = (64,))(dense_layer_2)

dense_layer_3     = Dense(32, activation = 'relu')(drop_out_layer_3)
drop_out_layer_4  = Dropout(0.5, input_shape = (32,))(dense_layer_3)

dense_layer_4     = Dense(10, activation = 'relu')(drop_out_layer_4)
drop_out_layer_5 = Dropout(0.5, input_shape = (10,))(dense_layer_4)

#-------------------------------------------------------------------------------
param = 1e-4

input_2 = Input(shape=(X_cat_train.shape[1],))
dense_layer_5       = Dense(10, input_dim=X_cat_train.shape[1], activation='relu', kernel_initializer='he_uniform', kernel_regularizer=l2(param),
                      kernel_constraint=unit_norm())(input_2)
drop_out_layer_6    = Dropout(0.2)(dense_layer_5)
batch_norm_layer_1  = BatchNormalization()(drop_out_layer_6)
dense_layer_6       = Dense(10, activation='relu', kernel_initializer='he_uniform', kernel_regularizer=l2(param), 
                            kernel_constraint=unit_norm())(batch_norm_layer_1)
drop_out_layer_7   = Dropout(0.5)(dense_layer_6)
batch_norm_layer_2 = BatchNormalization()(drop_out_layer_7)

concat_layer        = Concatenate()([drop_out_layer_5, batch_norm_layer_2])
dense_layer_7       = Dense(10, activation='relu')(concat_layer)
output  = Dense(5, activation='softmax')(dense_layer_7)
model   = Model(inputs=[input_1, input_2], outputs=output)

# compile the keras model
#opt = optimizers.Adamax(lr=0.01)
opt = SGD(lr=0.001, momentum=0.9)
model.compile(loss='categorical_crossentropy', optimizer=opt, metrics=['acc'])

print(model.summary())

Model: "model_2"
__________________________________________________________________________________________________
Layer (type)                    Output Shape         Param #     Connected to                     
==================================================================================================
input_3 (InputLayer)            [(None, 100)]        0                                            
__________________________________________________________________________________________________
embedding_1 (Embedding)         (None, 100, 200)     553000      input_3[0][0]                    
__________________________________________________________________________________________________
bidirectional_1 (Bidirectional) (None, 100, 256)     336896      embedding_1[0][0]                
__________________________________________________________________________________________________
global_max_pooling1d_1 (GlobalM (None, 256)          0           bidirectional_1[0][0]            
__________________________________________________________________________________________________
dropout_11 (Dropout)            (None, 256)          0           global_max_pooling1d_1[0][0]     
__________________________________________________________________________________________________
dense_19 (Dense)                (None, 128)          32896       dropout_11[0][0]                 
__________________________________________________________________________________________________
dropout_12 (Dropout)            (None, 128)          0           dense_19[0][0]                   
__________________________________________________________________________________________________
input_4 (InputLayer)            [(None, 12)]         0                                            
__________________________________________________________________________________________________
dense_20 (Dense)                (None, 64)           8256        dropout_12[0][0]                 
__________________________________________________________________________________________________
dense_23 (Dense)                (None, 10)           130         input_4[0][0]                    
__________________________________________________________________________________________________
dropout_13 (Dropout)            (None, 64)           0           dense_20[0][0]                   
__________________________________________________________________________________________________
dropout_16 (Dropout)            (None, 10)           0           dense_23[0][0]                   
__________________________________________________________________________________________________
dense_21 (Dense)                (None, 32)           2080        dropout_13[0][0]                 
__________________________________________________________________________________________________
batch_normalization_6 (BatchNor (None, 10)           40          dropout_16[0][0]                 
__________________________________________________________________________________________________
dropout_14 (Dropout)            (None, 32)           0           dense_21[0][0]                   
__________________________________________________________________________________________________
dense_24 (Dense)                (None, 10)           110         batch_normalization_6[0][0]      
__________________________________________________________________________________________________
dense_22 (Dense)                (None, 10)           330         dropout_14[0][0]                 
__________________________________________________________________________________________________
dropout_17 (Dropout)            (None, 10)           0           dense_24[0][0]                   
__________________________________________________________________________________________________
dropout_15 (Dropout)            (None, 10)           0           dense_22[0][0]                   
__________________________________________________________________________________________________
batch_normalization_7 (BatchNor (None, 10)           40          dropout_17[0][0]                 
__________________________________________________________________________________________________
concatenate (Concatenate)       (None, 20)           0           dropout_15[0][0]                 
                                                                 batch_normalization_7[0][0]      
__________________________________________________________________________________________________
dense_25 (Dense)                (None, 10)           210         concatenate[0][0]                
__________________________________________________________________________________________________
dense_26 (Dense)                (None, 5)            55          dense_25[0][0]                   
==================================================================================================
Total params: 934,043
Trainable params: 381,003
Non-trainable params: 553,040
__________________________________________________________________________________________________
None

# Use earlystopping
callback = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=7, min_delta=1E-3)
rlrp = ReduceLROnPlateau(monitor='val_loss', factor=0.0001, patience=5, min_delta=1E-4)

target_type = 'multi_label'
metrics = Metrics(validation_data=([X_text_train, X_cat_train], y_cat_train, target_type))

# fit the keras model on the dataset
training_history = model.fit([X_text_train, X_cat_train], y_cat_train, epochs=100, batch_size=8, verbose=1, validation_data=([X_text_test, X_cat_test], y_cat_test), callbacks=[rlrp, metrics])

Epoch 1/100
42/42 [==============================] - 14s 175ms/step - loss: 1.2883 - acc: 0.5824 - val_loss: 1.2806 - val_acc: 0.7381
Epoch 2/100
42/42 [==============================] - 5s 121ms/step - loss: 1.0993 - acc: 0.7330 - val_loss: 1.1707 - val_acc: 0.7381
Epoch 3/100
42/42 [==============================] - 6s 153ms/step - loss: 0.9987 - acc: 0.7401 - val_loss: 1.0878 - val_acc: 0.7381
Epoch 4/100
42/42 [==============================] - 5s 119ms/step - loss: 1.1155 - acc: 0.7029 - val_loss: 1.0573 - val_acc: 0.7381
Epoch 5/100
42/42 [==============================] - 6s 143ms/step - loss: 1.0526 - acc: 0.7398 - val_loss: 1.0369 - val_acc: 0.7381
Epoch 6/100
42/42 [==============================] - 5s 117ms/step - loss: 1.0358 - acc: 0.7424 - val_loss: 1.0295 - val_acc: 0.7381
Epoch 7/100
42/42 [==============================] - 7s 166ms/step - loss: 0.8943 - acc: 0.7843 - val_loss: 1.0264 - val_acc: 0.7381
Epoch 8/100
42/42 [==============================] - 6s 134ms/step - loss: 0.9843 - acc: 0.7423 - val_loss: 1.0300 - val_acc: 0.7381
Epoch 9/100
42/42 [==============================] - 4s 106ms/step - loss: 1.0080 - acc: 0.7416 - val_loss: 1.0266 - val_acc: 0.7381
Epoch 10/100
42/42 [==============================] - 5s 126ms/step - loss: 0.8780 - acc: 0.7867 - val_loss: 1.0202 - val_acc: 0.7381
Epoch 11/100
42/42 [==============================] - 5s 119ms/step - loss: 0.9912 - acc: 0.7192 - val_loss: 1.0118 - val_acc: 0.7381
Epoch 12/100
42/42 [==============================] - 5s 121ms/step - loss: 0.8995 - acc: 0.7695 - val_loss: 1.0040 - val_acc: 0.7381
Epoch 13/100
42/42 [==============================] - 5s 129ms/step - loss: 0.9242 - acc: 0.7617 - val_loss: 0.9961 - val_acc: 0.7381
Epoch 14/100
42/42 [==============================] - 5s 130ms/step - loss: 0.9033 - acc: 0.7502 - val_loss: 0.9847 - val_acc: 0.7381
Epoch 15/100
42/42 [==============================] - 5s 131ms/step - loss: 0.9511 - acc: 0.7488 - val_loss: 0.9810 - val_acc: 0.7381
Epoch 16/100
42/42 [==============================] - 6s 136ms/step - loss: 0.9139 - acc: 0.7615 - val_loss: 0.9783 - val_acc: 0.7381
Epoch 17/100
42/42 [==============================] - 5s 118ms/step - loss: 0.9572 - acc: 0.7338 - val_loss: 0.9813 - val_acc: 0.7381
Epoch 18/100
42/42 [==============================] - 6s 135ms/step - loss: 0.9479 - acc: 0.7345 - val_loss: 0.9820 - val_acc: 0.7381
Epoch 19/100
42/42 [==============================] - 5s 120ms/step - loss: 0.8623 - acc: 0.7851 - val_loss: 0.9742 - val_acc: 0.7381
Epoch 20/100
42/42 [==============================] - 5s 124ms/step - loss: 0.9881 - acc: 0.7210 - val_loss: 0.9731 - val_acc: 0.7381
Epoch 21/100
42/42 [==============================] - 5s 125ms/step - loss: 0.8770 - acc: 0.7669 - val_loss: 0.9697 - val_acc: 0.7381
Epoch 22/100
42/42 [==============================] - 5s 121ms/step - loss: 1.0772 - acc: 0.6944 - val_loss: 0.9722 - val_acc: 0.7381
Epoch 23/100
42/42 [==============================] - 5s 115ms/step - loss: 1.0284 - acc: 0.7093 - val_loss: 0.9677 - val_acc: 0.7381
Epoch 24/100
42/42 [==============================] - 5s 120ms/step - loss: 0.8860 - acc: 0.7660 - val_loss: 0.9615 - val_acc: 0.7381
Epoch 25/100
42/42 [==============================] - 5s 120ms/step - loss: 0.9106 - acc: 0.7509 - val_loss: 0.9654 - val_acc: 0.7381
Epoch 26/100
42/42 [==============================] - 5s 128ms/step - loss: 0.8340 - acc: 0.7719 - val_loss: 0.9622 - val_acc: 0.7381
Epoch 27/100
42/42 [==============================] - 5s 119ms/step - loss: 0.9397 - acc: 0.7271 - val_loss: 0.9647 - val_acc: 0.7381
Epoch 28/100
42/42 [==============================] - 5s 129ms/step - loss: 0.8626 - acc: 0.7628 - val_loss: 0.9582 - val_acc: 0.7381
Epoch 29/100
42/42 [==============================] - 5s 119ms/step - loss: 0.8901 - acc: 0.7406 - val_loss: 0.9569 - val_acc: 0.7381
Epoch 30/100
42/42 [==============================] - 5s 128ms/step - loss: 0.8799 - acc: 0.7652 - val_loss: 0.9535 - val_acc: 0.7381
Epoch 31/100
42/42 [==============================] - 7s 156ms/step - loss: 0.9428 - acc: 0.7407 - val_loss: 0.9565 - val_acc: 0.7381
Epoch 32/100
42/42 [==============================] - 6s 145ms/step - loss: 0.8583 - acc: 0.7582 - val_loss: 0.9547 - val_acc: 0.7381
Epoch 33/100
42/42 [==============================] - 6s 151ms/step - loss: 0.8441 - acc: 0.7760 - val_loss: 0.9545 - val_acc: 0.7381
Epoch 34/100
42/42 [==============================] - 6s 148ms/step - loss: 0.9580 - acc: 0.7256 - val_loss: 0.9566 - val_acc: 0.7381
Epoch 35/100
42/42 [==============================] - 6s 137ms/step - loss: 0.8426 - acc: 0.7737 - val_loss: 0.9575 - val_acc: 0.7381
Epoch 36/100
42/42 [==============================] - 5s 123ms/step - loss: 0.9064 - acc: 0.7470 - val_loss: 0.9597 - val_acc: 0.7381
Epoch 37/100
42/42 [==============================] - 5s 123ms/step - loss: 0.9563 - acc: 0.7260 - val_loss: 0.9599 - val_acc: 0.7381
Epoch 38/100
42/42 [==============================] - 5s 119ms/step - loss: 0.9608 - acc: 0.7300 - val_loss: 0.9594 - val_acc: 0.7381
Epoch 39/100
42/42 [==============================] - 5s 121ms/step - loss: 0.9089 - acc: 0.7425 - val_loss: 0.9592 - val_acc: 0.7381
Epoch 40/100
42/42 [==============================] - 5s 124ms/step - loss: 0.9077 - acc: 0.7408 - val_loss: 0.9592 - val_acc: 0.7381
Epoch 41/100
42/42 [==============================] - 5s 123ms/step - loss: 0.9469 - acc: 0.7275 - val_loss: 0.9587 - val_acc: 0.7381
Epoch 42/100
42/42 [==============================] - 5s 122ms/step - loss: 0.9870 - acc: 0.7093 - val_loss: 0.9585 - val_acc: 0.7381
Epoch 43/100
42/42 [==============================] - 5s 128ms/step - loss: 0.8359 - acc: 0.7632 - val_loss: 0.9585 - val_acc: 0.7381
Epoch 44/100
42/42 [==============================] - 5s 121ms/step - loss: 0.8443 - acc: 0.7625 - val_loss: 0.9593 - val_acc: 0.7381
Epoch 45/100
42/42 [==============================] - 6s 151ms/step - loss: 0.9086 - acc: 0.7388 - val_loss: 0.9587 - val_acc: 0.7381
Epoch 46/100
42/42 [==============================] - 7s 164ms/step - loss: 0.9003 - acc: 0.7490 - val_loss: 0.9591 - val_acc: 0.7381
Epoch 47/100
42/42 [==============================] - 5s 130ms/step - loss: 1.0354 - acc: 0.6878 - val_loss: 0.9590 - val_acc: 0.7381
Epoch 48/100
42/42 [==============================] - 5s 128ms/step - loss: 0.9445 - acc: 0.7411 - val_loss: 0.9596 - val_acc: 0.7381
Epoch 49/100
42/42 [==============================] - 6s 133ms/step - loss: 0.8935 - acc: 0.7465 - val_loss: 0.9605 - val_acc: 0.7381
Epoch 50/100
42/42 [==============================] - 5s 119ms/step - loss: 0.9476 - acc: 0.7343 - val_loss: 0.9596 - val_acc: 0.7381
Epoch 51/100
42/42 [==============================] - 5s 130ms/step - loss: 0.9938 - acc: 0.7049 - val_loss: 0.9592 - val_acc: 0.7381
Epoch 52/100
42/42 [==============================] - 6s 133ms/step - loss: 0.9300 - acc: 0.7419 - val_loss: 0.9590 - val_acc: 0.7381
Epoch 53/100
42/42 [==============================] - 5s 128ms/step - loss: 0.9268 - acc: 0.7465 - val_loss: 0.9600 - val_acc: 0.7381
Epoch 54/100
42/42 [==============================] - 5s 117ms/step - loss: 0.9332 - acc: 0.7156 - val_loss: 0.9592 - val_acc: 0.7381
Epoch 55/100
42/42 [==============================] - 5s 115ms/step - loss: 0.8984 - acc: 0.7365 - val_loss: 0.9600 - val_acc: 0.7381
Epoch 56/100
42/42 [==============================] - 5s 124ms/step - loss: 0.8665 - acc: 0.7642 - val_loss: 0.9602 - val_acc: 0.7381
Epoch 57/100
42/42 [==============================] - 6s 135ms/step - loss: 0.8915 - acc: 0.7513 - val_loss: 0.9596 - val_acc: 0.7381
Epoch 58/100
42/42 [==============================] - 5s 120ms/step - loss: 0.9510 - acc: 0.7199 - val_loss: 0.9600 - val_acc: 0.7381
Epoch 59/100
42/42 [==============================] - 6s 136ms/step - loss: 0.8834 - acc: 0.7640 - val_loss: 0.9602 - val_acc: 0.7381
Epoch 60/100
42/42 [==============================] - 5s 127ms/step - loss: 0.9335 - acc: 0.7187 - val_loss: 0.9595 - val_acc: 0.7381
Epoch 61/100
42/42 [==============================] - 6s 143ms/step - loss: 0.9648 - acc: 0.7131 - val_loss: 0.9596 - val_acc: 0.7381
Epoch 62/100
42/42 [==============================] - 6s 143ms/step - loss: 0.9236 - acc: 0.7353 - val_loss: 0.9591 - val_acc: 0.7381
Epoch 63/100
42/42 [==============================] - 6s 143ms/step - loss: 0.8773 - acc: 0.7597 - val_loss: 0.9588 - val_acc: 0.7381
Epoch 64/100
42/42 [==============================] - 5s 131ms/step - loss: 0.9760 - acc: 0.7075 - val_loss: 0.9590 - val_acc: 0.7381
Epoch 65/100
42/42 [==============================] - 7s 158ms/step - loss: 0.9837 - acc: 0.7116 - val_loss: 0.9590 - val_acc: 0.7381
Epoch 66/100
42/42 [==============================] - 7s 157ms/step - loss: 0.9601 - acc: 0.7220 - val_loss: 0.9596 - val_acc: 0.7381
Epoch 67/100
42/42 [==============================] - 5s 116ms/step - loss: 0.9032 - acc: 0.7422 - val_loss: 0.9592 - val_acc: 0.7381
Epoch 68/100
42/42 [==============================] - 5s 115ms/step - loss: 0.9880 - acc: 0.7083 - val_loss: 0.9591 - val_acc: 0.7381
Epoch 69/100
42/42 [==============================] - 6s 144ms/step - loss: 0.9173 - acc: 0.7367 - val_loss: 0.9593 - val_acc: 0.7381
Epoch 70/100
42/42 [==============================] - 5s 120ms/step - loss: 0.8855 - acc: 0.7496 - val_loss: 0.9596 - val_acc: 0.7381
Epoch 71/100
42/42 [==============================] - 7s 160ms/step - loss: 0.9971 - acc: 0.7223 - val_loss: 0.9594 - val_acc: 0.7381
Epoch 72/100
42/42 [==============================] - 6s 139ms/step - loss: 0.9749 - acc: 0.7138 - val_loss: 0.9601 - val_acc: 0.7381
Epoch 73/100
42/42 [==============================] - 5s 127ms/step - loss: 0.8935 - acc: 0.7683 - val_loss: 0.9596 - val_acc: 0.7381
Epoch 74/100
42/42 [==============================] - 5s 119ms/step - loss: 0.8660 - acc: 0.7555 - val_loss: 0.9587 - val_acc: 0.7381
Epoch 75/100
42/42 [==============================] - 5s 117ms/step - loss: 0.8729 - acc: 0.7509 - val_loss: 0.9587 - val_acc: 0.7381
Epoch 76/100
42/42 [==============================] - 6s 146ms/step - loss: 0.9286 - acc: 0.7371 - val_loss: 0.9593 - val_acc: 0.7381
Epoch 77/100
42/42 [==============================] - 6s 146ms/step - loss: 0.9207 - acc: 0.7435 - val_loss: 0.9586 - val_acc: 0.7381
Epoch 78/100
42/42 [==============================] - 6s 137ms/step - loss: 0.9228 - acc: 0.7517 - val_loss: 0.9594 - val_acc: 0.7381
Epoch 79/100
42/42 [==============================] - 5s 122ms/step - loss: 0.8885 - acc: 0.7537 - val_loss: 0.9589 - val_acc: 0.7381
Epoch 80/100
42/42 [==============================] - 6s 134ms/step - loss: 0.9347 - acc: 0.7305 - val_loss: 0.9593 - val_acc: 0.7381
Epoch 81/100
42/42 [==============================] - 7s 172ms/step - loss: 0.9419 - acc: 0.7464 - val_loss: 0.9589 - val_acc: 0.7381
Epoch 82/100
42/42 [==============================] - 6s 155ms/step - loss: 0.9679 - acc: 0.7456 - val_loss: 0.9601 - val_acc: 0.7381
Epoch 83/100
42/42 [==============================] - 6s 138ms/step - loss: 0.9394 - acc: 0.7247 - val_loss: 0.9609 - val_acc: 0.7381
Epoch 84/100
42/42 [==============================] - 7s 160ms/step - loss: 0.8292 - acc: 0.7740 - val_loss: 0.9603 - val_acc: 0.7381
Epoch 85/100
42/42 [==============================] - 6s 141ms/step - loss: 0.8852 - acc: 0.7485 - val_loss: 0.9590 - val_acc: 0.7381
Epoch 86/100
42/42 [==============================] - 6s 136ms/step - loss: 0.8813 - acc: 0.7424 - val_loss: 0.9593 - val_acc: 0.7381
Epoch 87/100
42/42 [==============================] - 6s 142ms/step - loss: 0.9017 - acc: 0.7330 - val_loss: 0.9582 - val_acc: 0.7381
Epoch 88/100
42/42 [==============================] - 5s 124ms/step - loss: 0.9421 - acc: 0.7352 - val_loss: 0.9591 - val_acc: 0.7381
Epoch 89/100
42/42 [==============================] - 5s 129ms/step - loss: 0.8430 - acc: 0.7732 - val_loss: 0.9591 - val_acc: 0.7381
Epoch 90/100
42/42 [==============================] - 6s 145ms/step - loss: 0.9214 - acc: 0.7358 - val_loss: 0.9588 - val_acc: 0.7381
Epoch 91/100
42/42 [==============================] - 5s 131ms/step - loss: 0.9131 - acc: 0.7492 - val_loss: 0.9587 - val_acc: 0.7381
Epoch 92/100
42/42 [==============================] - 5s 123ms/step - loss: 0.9145 - acc: 0.7384 - val_loss: 0.9583 - val_acc: 0.7381
Epoch 93/100
42/42 [==============================] - 6s 138ms/step - loss: 0.8787 - acc: 0.7564 - val_loss: 0.9592 - val_acc: 0.7381
Epoch 94/100
42/42 [==============================] - 6s 140ms/step - loss: 0.8922 - acc: 0.7382 - val_loss: 0.9600 - val_acc: 0.7381
Epoch 95/100
42/42 [==============================] - 5s 108ms/step - loss: 0.9269 - acc: 0.7271 - val_loss: 0.9591 - val_acc: 0.7381
Epoch 96/100
42/42 [==============================] - 7s 168ms/step - loss: 0.8060 - acc: 0.7866 - val_loss: 0.9598 - val_acc: 0.7381
Epoch 97/100
42/42 [==============================] - 6s 131ms/step - loss: 0.9641 - acc: 0.7155 - val_loss: 0.9598 - val_acc: 0.7381
Epoch 98/100
42/42 [==============================] - 5s 128ms/step - loss: 0.9368 - acc: 0.7255 - val_loss: 0.9598 - val_acc: 0.7381
Epoch 99/100
42/42 [==============================] - 6s 134ms/step - loss: 0.9011 - acc: 0.7504 - val_loss: 0.9597 - val_acc: 0.7381
Epoch 100/100
42/42 [==============================] - 7s 169ms/step - loss: 1.0093 - acc: 0.7125 - val_loss: 0.9604 - val_acc: 0.7381

# evaluate the keras model
_, train_accuracy = model.evaluate([X_text_train, X_cat_train], y_cat_train, batch_size=8, verbose=0)
_, test_accuracy = model.evaluate([X_text_test, X_cat_test], y_cat_test, batch_size=8, verbose=0)

print(color.GREEN + 'Train accuracy: ' + color.END + str(train_accuracy*100))
print(color.GREEN + 'Test accuracy: ' + color.END + str(test_accuracy*100))

Train accuracy: 73.95209670066833
Test accuracy: 73.8095223903656

accuracy, precision, recall, f1 = get_classification_metrics(model, [X_text_test, X_cat_test], y_cat_test, target_type)
print(color.BLUE + 'Accuracy: ' + color.END + str(accuracy))
print(color.BLUE + 'Precision: ' + color.END + str(precision))
print(color.BLUE + 'Recall: ' + color.END + str(recall))
print(color.BLUE + 'F1 score: ' + color.END + str(f1))

Accuracy: 0.7380952380952381
Precision: 0.7380952380952381
Recall: 0.7380952380952381
F1 score: 0.7380952380952381

epochs = range(len(training_history.history['loss'])) # Get number of epochs

# plot loss learning curves
plt.plot(epochs, training_history.history['loss'], label = 'train')
plt.plot(epochs, training_history.history['val_loss'], label = 'test')
plt.legend(loc = 'upper right')
plt.title ('Training and validation loss')

Text(0.5,1,'Training and validation loss')

# plot accuracy learning curves
plt.plot(epochs, training_history.history['acc'], label = 'train')
plt.plot(epochs, training_history.history['val_acc'], label = 'test')
plt.legend(loc = 'upper right')
plt.title ('Training and validation accuracy')

Text(0.5,1,'Training and validation accuracy')

from io import StringIO
def chatbot_response(msg):
    op=""
    final_model = load_model("finalized_keras_model_v0.h5")
    s="desc"
    s = s+"\n"+msg
    StringData = StringIO(s)
    # let's read the data using the Pandas
    # read_csv() function
    df = pd.read_csv(StringData, sep =";")
    df_op = get_cleaned_desc_wo_print(df, 'desc', 'cleaned_desc')
    tokenizer = Tokenizer(num_words=5000)
    tokenizer.fit_on_texts(df_op)
    df_op = tokenizer.texts_to_sequences(df_op)
    vocab_size = len(tokenizer.word_index) + 1
    maxlen = 100
    df_op = pad_sequences(df_op, padding='post', maxlen=maxlen)
    seq_predictions=final_model.predict(df_op[0]) 
    seq_predictions=seq_predictions[0]  # transformation to get (n,)
    seq_predictions = list(map(lambda x: 0 if x<0.5 else 1, seq_predictions))
    max_index = seq_predictions.index(max(seq_predictions)) + 1 
    if max_index==1:
        op = 'The accident Level is Minor'
    elif max_index==2:
        op = 'The accident Level is Moderate'
    elif max_index==3:
        op = 'The accident Level is Severe'
    elif max_index==4:
        op = 'The accident Level is Critical'
    elif max_index==5:
        op = 'The accident Level is Fatal' 
    return op

#Creating GUI with tkinter
import tkinter
from tkinter import *
def send():
    msg = EntryBox.get("1.0",'end-1c').strip()
    EntryBox.delete("0.0",END)
    if (msg.lower() == 'hi') or (msg.lower() == 'hello'):
        ChatLog.config(state=NORMAL)
        ChatLog.insert(END, "You: " + msg + '\n\n')
        ChatLog.config(foreground="#442265", font=("Verdana", 12 ))
        ChatLog.insert(END, "Bot: Hello! Kindly put in the description of the accident" + '\n\n')
        ChatLog.config(state=DISABLED)
        ChatLog.yview(END)
    elif msg != '':
        ChatLog.config(state=NORMAL)
        ChatLog.insert(END, "You: " + msg + '\n\n')
        ChatLog.config(foreground="#442265", font=("Verdana", 12 ))
        res = chatbot_response(msg)
        ChatLog.insert(END, "Bot: " + res + '\n\n')
        ChatLog.config(state=DISABLED)
        ChatLog.yview(END)
base = Tk()
base.title("Accident Level Prediction Bot")
base.geometry("400x500")
base.resizable(width=FALSE, height=FALSE)
#Create Chat window
ChatLog = Text(base, bd=0, bg="white", height="8", width="50", font="Arial",)
ChatLog.config(state=DISABLED)
#Bind scrollbar to Chat window
scrollbar = Scrollbar(base, command=ChatLog.yview, cursor="heart")
ChatLog['yscrollcommand'] = scrollbar.set
#Create Button to send message
SendButton = Button(base, font=("Verdana",12,'bold'), text="Send", width="12", height=5,
                    bd=0, bg="#32de97", activebackground="#3c9d9b",fg='#ffffff',
                    command= send )
#Create the box to enter message
EntryBox = Text(base, bd=0, bg="white",width="29", height="5", font="Arial")
#EntryBox.bind("<Return>", send)
#Place all components on the screen
scrollbar.place(x=376,y=6, height=386)
ChatLog.place(x=6,y=6, height=386, width=370)
EntryBox.place(x=128, y=401, height=90, width=265)
SendButton.place(x=6, y=401, height=90)
base.mainloop()

	Unnamed: 0	Data	Countries	Local	Industry Sector	Accident Level	Potential Accident Level	Genre	Employee or Third Party	Critical Risk	Description
0	0	2016-01-01 00:00:00	Country_01	Local_01	Mining	I	IV	Male	Third Party	Pressed	While removing the drill rod of the Jumbo 08 f...
1	1	2016-01-02 00:00:00	Country_02	Local_02	Mining	I	IV	Male	Employee	Pressurized Systems	During the activation of a sodium sulphide pum...
2	2	2016-01-06 00:00:00	Country_01	Local_03	Mining	I	III	Male	Third Party (Remote)	Manual Tools	In the sub-station MILPO located at level +170...
3	3	2016-01-08 00:00:00	Country_01	Local_04	Mining	I	I	Male	Third Party	Others	Being 9:45 am. approximately in the Nv. 1880 C...
4	4	2016-01-10 00:00:00	Country_01	Local_04	Mining	IV	IV	Male	Third Party	Others	Approximately at 11:45 a.m. in circumstances t...

	0	NUnique
Unnamed: 0	int64	425
Data	object	287
Countries	object	3
Local	object	12
Industry Sector	object	3
Accident Level	object	5
Potential Accident Level	object	6
Genre	object	2
Employee or Third Party	object	3
Critical Risk	object	33
Description	object	411

	Data	Countries	Local	Industry Sector	Accident Level	Potential Accident Level	Genre	Employee or Third Party	Critical Risk	Description
count	425	425	425	425	425	425	425	425	425	425
unique	287	3	12	3	5	6	2	3	33	411
top	2017-02-08 00:00:00	Country_01	Local_03	Mining	I	IV	Male	Third Party	Others	On 02/03/17 during the soil sampling in the re...
freq	6	251	90	241	316	143	403	189	232	3

	date	country	local	industry_sector	accident_level	potential_accident_level	gender	emp_type	critical_risk	description
77	2016-04-01 00:00:00	Country_01	Local_01	Mining	I	V	Male	Third Party (Remote)	Others	In circumstances that two workers of the Abrat...
262	2016-12-01 00:00:00	Country_01	Local_03	Mining	I	IV	Male	Employee	Others	During the activity of chuteo of ore in hopper...
303	2017-01-21 00:00:00	Country_02	Local_02	Mining	I	I	Male	Third Party (Remote)	Others	Employees engaged in the removal of material f...
345	2017-03-02 00:00:00	Country_03	Local_10	Others	I	I	Male	Third Party	Venomous Animals	On 02/03/17 during the soil sampling in the re...
346	2017-03-02 00:00:00	Country_03	Local_10	Others	I	I	Male	Third Party	Venomous Animals	On 02/03/17 during the soil sampling in the re...
355	2017-03-15 00:00:00	Country_03	Local_10	Others	I	I	Male	Third Party	Venomous Animals	Team of the VMS Project performed soil collect...
397	2017-05-23 00:00:00	Country_01	Local_04	Mining	I	IV	Male	Third Party	Projection of fragments	In moments when the 02 collaborators carried o...

	date	country	local	industry_sector	accident_level	potential_accident_level	gender	emp_type	critical_risk	description	year	month	day	weekday	week_of_year
0	2016-01-01	Country_01	Local_01	Mining	I	IV	Male	Third Party	Pressed	While removing the drill rod of the Jumbo 08 f...	2016	1	1	Friday	53
1	2016-01-02	Country_02	Local_02	Mining	I	IV	Male	Employee	Pressurized Systems	During the activation of a sodium sulphide pum...	2016	1	2	Saturday	53
2	2016-01-06	Country_01	Local_03	Mining	I	III	Male	Third Party (Remote)	Manual Tools	In the sub-station MILPO located at level +170...	2016	1	6	Wednesday	1
3	2016-01-08	Country_01	Local_04	Mining	I	I	Male	Third Party	Others	Being 9:45 am. approximately in the Nv. 1880 C...	2016	1	8	Friday	1
4	2016-01-10	Country_01	Local_04	Mining	IV	IV	Male	Third Party	Others	Approximately at 11:45 a.m. in circumstances t...	2016	1	10	Sunday	1

	tfidf_area	tfidf_causing	tfidf_hand	tfidf_injury	tfidf_left	tfidf_causing injury	tfidf_left foot
0	0.00	0.00	0.00	0.00	0.00	0.00	0.00
1	1.00	0.00	0.00	0.00	0.00	0.00	0.00
2	0.55	0.38	0.41	0.46	0.40	0.58	0.81

	year	month	day	week_of_year	season	weekday	country	local	gender	industry_sector	emp_type	critical_risk	tfidf_activity	tfidf_employee	tfidf_hand	tfidf_left	tfidf_right	tfidf_left hand	tfidf_right hand
202	-0.70	1.17	-1.67	1.00	1.73	0.08	0.75	0.10	0.24	-1.29	1.91	-0.29	0.00	0.00	0.71	0.00	0.70	0.00	1.00
258	-0.70	1.80	1.56	1.93	1.73	-1.00	0.75	1.17	0.24	-1.29	-1.05	0.01	0.45	0.37	0.73	0.36	0.00	1.00	0.00
6	-0.70	-1.36	0.06	-1.38	-1.22	1.15	0.75	0.10	0.24	-1.29	-1.05	-0.89	0.00	0.00	0.84	0.55	0.00	1.00	0.00

	Method	Train Accuracy	Test Accuracy	Precision	Recall	F1-Score	Multi-Class Logloss
1	LogisticRegression	0.74	0.73	0.56	0.73	0.63	0.88
2	RidgeClassifier	0.74	0.74	0.56	0.74	0.64	1.00
3	KNeighborsClassifier	0.75	0.70	0.57	0.70	0.63	6.48
4	SVC	0.74	0.74	0.54	0.74	0.63	0.90
5	DecisionTreeClassifier	0.99	0.64	0.66	0.64	0.65	12.34
6	RandomForestClassifier	0.95	0.73	0.56	0.73	0.63	3.07
7	BaggingClassifier	0.96	0.73	0.62	0.73	0.64	2.39
8	ExtraTreesClassifier	0.99	0.74	0.55	0.74	0.63	0.82
9	AdaBoostClassifier	0.73	0.74	0.67	0.74	0.66	1.23
10	GradientBoostingClassifier	0.94	0.68	0.55	0.68	0.61	0.96
11	LGBMClassifier	0.99	0.74	0.64	0.74	0.68	1.14
12	XGBClassifier	0.82	0.74	0.59	0.74	0.65	0.85

	Method	Train Accuracy	Test Accuracy	Precision	Recall	F1-Score	Multi-Class Logloss
1	LogisticRegression	0.74	0.26	0.68	0.26	0.35	1.77
2	RidgeClassifier	0.63	0.24	0.67	0.24	0.29	1.00
3	KNeighborsClassifier	0.95	0.50	0.63	0.50	0.55	11.75
4	SVC	0.97	0.56	0.59	0.56	0.58	1.31
5	DecisionTreeClassifier	1.00	0.61	0.65	0.61	0.63	13.57
6	RandomForestClassifier	1.00	0.74	0.64	0.74	0.67	3.09
7	BaggingClassifier	1.00	0.71	0.64	0.71	0.68	2.02
8	ExtraTreesClassifier	1.00	0.74	0.61	0.74	0.65	1.65
9	AdaBoostClassifier	0.40	0.61	0.59	0.61	0.60	1.30
10	GradientBoostingClassifier	0.97	0.57	0.63	0.57	0.60	1.06
11	LGBMClassifier	1.00	0.73	0.57	0.73	0.64	1.72
12	XGBClassifier	1.00	0.68	0.63	0.68	0.64	0.95

	Method	Train Accuracy	Test Accuracy	Precision	Recall	F1-Score	Multi-Class Logloss
1	LogisticRegression	0.77	0.26	0.57	0.26	0.33	1.92
2	RidgeClassifier	0.71	0.24	0.61	0.24	0.31	1.00
3	KNeighborsClassifier	0.92	0.32	0.55	0.32	0.38	16.38
4	SVC	0.95	0.43	0.54	0.43	0.47	1.51
5	DecisionTreeClassifier	1.00	0.46	0.60	0.46	0.52	18.50
6	RandomForestClassifier	0.99	0.57	0.59	0.57	0.58	3.67
7	BaggingClassifier	1.00	0.57	0.58	0.57	0.57	2.98
8	ExtraTreesClassifier	1.00	0.56	0.58	0.56	0.57	1.14
9	AdaBoostClassifier	0.58	0.55	0.56	0.55	0.55	1.33
10	GradientBoostingClassifier	0.96	0.55	0.58	0.55	0.56	1.20
11	LGBMClassifier	1.00	0.65	0.59	0.65	0.62	1.72
12	XGBClassifier	0.99	0.58	0.58	0.58	0.58	1.16